This is not an essay. Normal service resumes in a fortnight.

For those who want to get Back to Marketing, I have five B2B Marketing and Marketing & AI courses that are now live. And I need your feedback!

Here is my offer:

The first twenty people to claim a place get twelve months of FP Collectiv Membership, which includes all current and future courses at no cost. Five courses, which come to US$1,095 bought separately, plus access to all new courses published during the year.

In return I want one brief survey when you finish your first course. Your feedback matters. If a module is weak, I would rather hear it from you now, so that I can make it better.

The courses cover B2B Marketing Fundamentals, Marketing Media and Measurement, and the AI track from Marketing & AI Foundations through to Productivity.

Twenty places only.

Claim a place

If it is not for you, the essays here stay free and fortnightly. Subscribe to the Back to Marketing essays below.

AI didn’t make marketing mediocre. Marketing was already mediocre.

AI simply made mediocre marketing almost free.

That distinction decides what you do next. If AI caused the problem, the fix is a policy about AI. If AI removed the cost of a problem the industry already had, then the fix is a much harder conversation about what your marketing was ever actually good at.

Start with the number that gives the game away. Eighty seven per cent of B2B marketers say AI has made them more productive. Fewer than two in five say their content is performing better. Both figures come from the Content Marketing Institute and MarketingProfs, surveying 1,015 B2B marketers in mid 2025. The same study found 58 per cent believe their content quality has improved.

So quality is up, output is up, productivity is up, and results are flat.

That is not a paradox. It is what happens when everyone gets better at the same time. Improvement that is universally available stops being improvement and becomes the new floor. AI made competence abundant. And once competence is widely available, it stops commanding an advantage.

Every marketing advantage dies the same way

Think about what has genuinely delivered advantage in marketing over the last thirty years, and when each one stopped.

Having a website. Understanding how search worked. Buying media efficiently. Owning a marketing automation stack. Publishing consistently. Running paid social before your category noticed it existed.

Every one of those was a real edge. Every one is now hygiene. And each died the same death, which had nothing to do with the capability getting worse. It got easy.

Advantage is not a property of a capability. It is a property of that capability’s scarcity.

The industry keeps missing this, because it keeps mistaking the capability for the advantage and then acting surprised when the returns evaporate. Media buying threw off outsized returns while it required specialist knowledge and relationships. Programmatic made competent buying available to anyone with a budget, and the advantage moved on. Search threw off outsized returns while most companies did not understand it, and then everyone hired for it. Marketing automation was a moat until the moat came in a subscription.

Nothing that can be installed is a moat.

Notice too that the half-life keeps shortening. Websites took years to travel from advantage to table stakes. Marketing automation moved faster. Content marketing faster again. Generative AI has compressed the cycle dramatically.

By the time the Content Marketing Institute fielded that survey in mid 2025, 89 per cent of B2B marketers were already using AI to generate or optimise copy. A capability sitting at 89 per cent adoption is not an advantage. It is a baseline. If your team is still writing the business case for AI adoption, the thing you are building the case for stopped differentiating anybody some time ago.

This is the honest answer to “how do we get an edge from AI”, and nobody wants to hear it. You don’t. Not from adoption. Adoption is now the cost of staying in the game, and the edge has already moved to whatever AI left scarce.

Hold that question. The rest of this follows from it.

The mediocrity came first

Before blaming the tools, look at what marketing was producing when the tools were expensive.

System1, working with LinkedIn’s B2B Institute, tested 1,600 B2B ads, mostly television, with around six million people globally over four years. Seventy five per cent scored one star or less on System1’s emotional response measure, the level that contributes effectively nothing to long term market share growth. Not one of the 1,600 reached five stars.

That was 2021. That was work with budget behind it, agencies attached, and senior sign off. It still landed as noise.

System1’s later research with eatbigfish and Peter Field put a figure on the flatness. In the UK market, 60 per cent of responses to B2B television advertising registered as emotionally neutral, against 52 per cent for consumer advertising. B2B is the duller category by measurement, not by reputation. And at the same level of excess share of voice, non dull advertising delivered around 1.3 points of annualised share gain while extremely dull advertising delivered around 0.1. An order of magnitude difference in growth, decided by whether the work registered at all.

Peter Field’s work for the IPA traced the same decline across nearly 600 case studies. Creatively awarded campaigns were roughly twelve times as efficient as non awarded campaigns in the twelve years to 2008. By 2018 that advantage had fallen below four times. Over the same period, the share of awarded campaigns that were short term rose from under 5 per cent to almost 40 per cent.

None of that required a language model. The industry manufactured its own mediocrity through short termism, measurement pressure and the steady substitution of activation for brand, which is the argument the previous essay in this series makes in full.

What generative AI changed is the price. Producing forgettable work used to cost a budget, a team and a quarter. Now it costs a prompt.

AI lowered the cost of being forgettable, and volume responded accordingly.

Average is now infinitely scalable

There is a mechanism behind the convergence, and it is better documented than most of the commentary suggests.

Anil Doshi at UCL and Oliver Hauser at the University of Exeter randomised 293 writers across three conditions: no AI, one AI generated idea, and five. Six hundred independent evaluators assessed the results, published in Science Advances in July 2024.

Individual work got better. Novelty rose 5.4 per cent with one AI idea and 8.1 per cent with five. Usefulness rose 3.7 and 9.0 per cent.

Collective similarity between the stories rose too, by 10.7 per cent of the measured range with a single AI idea.

Every writer improved. The body of work became more alike. For one marketer that reads as a win, because your draft beats the one you would have written alone. For a category it reads as convergence, because every competitor’s draft improved in the same direction at the same time.

There is a real complication worth knowing. A University of Michigan study with more than 800 participants across 40 countries found the opposite effect, with high AI exposure increasing collective idea diversity. The designs differ in one respect. Doshi and Hauser gave people AI output to work from. Michigan used AI to surface ideas participants would not otherwise have reached.

Substitution converges. Expansion diverges. Which one you get is a decision about process, not a property of the technology, and most teams have never made that decision consciously.

The volume side is now measurable. Graphite, sampling 43,000 URLs from Common Crawl, estimates that roughly half of newly published web articles are primarily AI generated. Its detector has published false-positive and false-negative rates, so treat the exact percentage cautiously. The important finding is not whether AI has technically crossed 50 per cent. It is that human and primarily AI-generated articles are now being produced at roughly comparable volumes.

What happens to that content is the more interesting finding. Across 31,493 keywords, of the articles ranking on the first two pages of Google results, 86 per cent were human written. Only 7 per cent of top position results were AI generated, half the baseline rate. Citations inside AI answer engines split roughly 82 per cent human to 18 per cent AI.

Production share and visibility share have come apart.

Meanwhile the surface being competed for is shrinking. Pew Research Center tracked 68,879 Google searches from a panel of 900 US adults. When an AI summary appeared, 8 per cent of visits produced a click on a traditional result, against 15 per cent when no summary appeared. One per cent clicked a link inside the summary itself. Ahrefs, comparing 150,000 keywords with AI Overviews against 150,000 without, reported a 58 per cent average drop in click through rate for top ranking pages, up from 34.5 per cent eight months earlier. SparkToro, using Similarweb clickstream data, put the share of US Google searches ending without any click at 68 per cent for early 2026.

More competent content, produced faster, competing for a materially smaller surface.

And the audience is not neutral about it. The Nuremberg Institute for Market Decisions ran controlled experiments alongside a 3,000 person survey across the US, UK and Germany. Identical advertising rated lower on emotional impact when labelled AI generated than when labelled human made, and only a quarter of respondents thought they could recognise AI content in the first place.

Which makes disclosure the wrong thing to be anxious about. Undifferentiated work fails on its own terms, long before anyone forms a view about how it was made.

Differentiation was never the mechanism

Here the industry’s own framing gets in the way, and it is worth clearing out.

Almost every article about B2B sameness quotes a statistic along the lines of “86 per cent of B2B buyers see no real difference between suppliers”, usually credited to CEB and Google. We went looking for the primary source. It does not exist. The figure circulates through agency blogs and conference decks citing each other, and it has been doing so for a decade.

Here is what is documented. Gartner surveyed more than 1,100 customers in late 2020 and found 64 per cent could not distinguish one B2B brand’s digital experience from a competitor’s. Gartner also names perceptions of difference between supplier offerings as one of three drivers of buyer confidence, and its earlier work found that buyers overwhelmed by high quality information were 153 per cent more likely to settle for a smaller, less ambitious purchase than they originally planned.

Sit with that one. The information was good. Volume alone was enough to shrink the deal.

Now the deeper point, which cuts against how most B2B teams think about this entirely.

In a 2007 paper in the Australasian Marketing Journal, Jenni Romaniuk, Byron Sharp and Andrew Ehrenberg showed a low level of perceived differentiation between competing brands across many categories and two countries. Buyers largely did not see meaningful differences. They bought anyway. The authors’ conclusion was to stop putting perceived differentiation at the centre of brand strategy and put distinctiveness there instead: unique associations that make a brand easily identifiable.

If distinctiveness rather than differentiation is the engine, then “everything sounds the same” describes a memory failure, not a positioning one. The question is identifiability. Does anything about your work make it retrievable later?

Retrievable when, though, is the part that gets skipped. John Dawes’s work at Ehrenberg-Bass with LinkedIn’s B2B Institute is the useful frame here, shorthanded as the 95-5 rule: up to 95 per cent of business buyers are not in market at any given moment. Dawes presents that as a heuristic rather than a measured constant, so use it as a way of thinking and not as a number in a board deck.

The measured version comes from 6sense’s B2B Buyer Experience Report for 2025, drawing on nearly 4,000 responses. Ninety five per cent of winning vendors were already on the buyer’s day one shortlist. Around four in five deals went to the vendor preferred before any seller contact. Buyers reached first contact roughly 61 per cent of the way through their process, earlier than the 69 per cent recorded the year before. Companion research put 85 per cent of buyers as having prior experience with the vendor they selected.

That is a statement about winners rather than about your odds, so do not turn it into a conversion rate. What it establishes is enough: by the time you reach a competitive evaluation, most of the outcome is already sitting in the buyer’s memory of you.

Which is where Jenni Romaniuk’s category entry points come in. Buyers retrieve brands by situation: the problem, the moment, the trigger. Attach yourself to a handful of the situations that matter and you get recalled when one occurs. Publish competent, unmemorable content attached to nothing in particular and you do not.

So the real cost of sameness lands well before the comparison. You were never in the set to be compared.

What abundance leaves scarce

Back to the question the second section left open. Every capability that becomes abundant stops conferring advantage. So what has survived every one of these cycles without becoming abundant?

Two things, and they are not the same thing, though the industry uses them interchangeably.

Judgement determines what should be done. Taste determines how it should be expressed.

Judgement is the commercial call: which market, which problem, which trade off, what to stop doing. The expressive call comes later and answers a different question, about what makes the work land, what to cut, and what deserves to exist at all.

Neither has ever been available in a subscription. You cannot install taste, and no prompt can make the final judgement for you.

The evidence for the judgement premium is strong and getting stronger. Fabrizio Dell’Acqua, Ethan Mollick, Karim Lakhani and colleagues ran an experiment with 758 consultants at Boston Consulting Group, published in Organization Science in 2025. On tasks inside AI’s capability frontier, consultants with AI access completed 12.2 per cent more tasks, 25.1 per cent faster, at higher quality. On tasks outside that frontier, consultants with AI access were 19 percentage points less likely to produce a correct solution than those working without it.

Same tool, same people. The variable was whether the task sat inside the model’s competence or outside it, and nobody hands you a map of where that line falls. Finding it is a judgement call, and the study prices what getting it wrong costs.

AI does not reduce the need for judgement. It raises the price of not having any, because a confident wrong answer produced in four seconds travels a great deal further than a slow one.

The labour market is already repricing this. Stanford’s Digital Economy Lab, using ADP payroll data covering millions of US workers, found employment among 22 to 25 year olds in AI exposed occupations now sits 19 per cent below where it would be had it kept pace with less exposed peers. Experienced workers show no comparable gap. When this series covered that research in the second essay, the figure was 16 per cent. Four months later it is 19, and the researchers note the divergence has widened steadily since they first documented it.

The finding underneath the headline is the one to carry. Declines concentrate where AI substitutes for human tasks. Where it complements, employment is flat or rising, particularly for experienced workers.

Now taste, which the industry treats as a soft word for something with a very hard price tag.

Go back to the System1, eatbigfish and Field research on dullness. In their UK B2B sample, of roughly £103 million in advertising spend, £53 million sat in the extremely dull quartile. More than half of the money went into the worst quarter of the work.

That is not a budget problem. Every one of those campaigns was funded, approved, and produced to a professional standard. Somebody signed it off. What was missing was the discernment to look at competent work and say it is not good enough to be remembered, and the standing to say it before the money went out the door.

Taste is expensive to lack and impossible to buy. Anu Atluru’s definition is the most useful one available: taste is discernment expressed. In a world of scarcity we treasure tools. In a world of abundance we treasure taste.

Marketing is now firmly in the second world, and most of its processes were built for the first.

There is a complication the profession has barely started to face. Judgement and taste are built by doing the work: running the campaign, defending the budget, watching a decision fail and understanding why. David Duncan argued in Harvard Business Review in February 2026 that AI is absorbing the repetitive early career tasks which historically produced professional discernment. Put that next to the Stanford data and the shape is uncomfortable. Fewer entry level roles, and less formative work inside the roles that remain, squeezes the pipeline that produces the one capability the market is now paying a premium for.

The scarce asset is getting scarcer. Which is good news for anyone who already has it, and a serious problem for the profession.

Five decisions worth making now

Diagnosis without a decision is commentary. Here is what changes if you accept the argument.

1. Stop looking for advantage in adoption.

A capability your competitors can buy this quarter is not, by itself, a strategy. Ask a harder question about anything you are about to invest in: how long before this is available to everyone in my category, and what happens to my position then. If the answer is under two years, you are buying hygiene and should budget for it as such rather than pitching it as an edge.

2. Measure retrieval, not output.

Volume metrics reward the one thing AI made free. If your reporting is built on pieces published, keywords ranked and sessions delivered, you have instrumented the part of the system that no longer differentiates you, in an environment where the click surface is contracting.

Add a memory side measure: whether your brand is retrieved for the buying situations that matter, tracked over time. The trade off is real. It moves slowly and is much harder to defend in a monthly review than a traffic chart. Make the case before you need it, not during a budget cut.

3. Decide your category entry points before you brief anything.

Identify the buying situations you want to be retrieved for and keep the list short enough to be memorable. Filter each candidate on three tests: is the association credible for you, is it competitive in the category, and is it common enough among buyers to be worth owning.

Everything you publish attaches to one of them or it does not get made. If your content calendar is currently a list of topics, converting it to a list of situations is the highest leverage change available to you this quarter.

4. Move AI upstream of the draft.

The homogenisation pattern comes from taking AI output as a starting point and polishing it. The diversity pattern comes from using AI to reach options you would not have reached alone. Same tools, different placement in the process.

Use it to widen the option set, pressure test a position, argue against your own case, find the evidence that contradicts you. Then reject the first plausible output as a matter of routine. This is slower per asset, and it only works if someone in the room has the standing to say no. Both of those are the point.

5. Protect the apprenticeship in your own team.

If the Stanford pattern reaches marketing and entry level work keeps thinning, an individual team develops a judgement gap long before the profession does. Hand juniors decisions rather than tasks. Let them own a call, defend it, and be wrong occasionally with support.

The cost is short term output. Weigh it against a team that can operate the tools and cannot tell when the tools are wrong.

What this comes down to

Every marketing advantage of the last thirty years died by becoming abundant. Websites, search, media buying, automation, content. AI is running the same cycle at four times the speed, and adoption crossed into hygiene while most teams were still building the business case.

Competent execution is no longer where sustainable advantage lives. It cannot be, because competent execution is getting cheaper and more widely available by the day.

What is left is the judgement to decide what deserves to exist, and the taste to make it worth remembering. Neither can be bought, installed, subscribed to or prompted. Both are built slowly, by people doing the work, which is why the supply is tightening at the exact moment the price is going up.

That is an uncomfortable place for the profession. It is an extraordinary opportunity for anyone willing to build the thing that does not scale.

Evidence informs. Judgement decides.

This is the fourth essay in a series on what marketing becomes when execution stops being the constraint. Subscribe to get the next one, which goes deep on the system that actually produces growth and where AI sits inside it.

If you want to build this properly rather than read about it, FP Collectiv’s courses take the same evidence led approach across B2B marketing fundamentals and marketing with AI.

Sources

Content Marketing Institute and MarketingProfs, B2B Content and Marketing Trends: Insights for 2026, published October 2025. Survey of 1,015 B2B marketers conducted June to August 2025. Findings used: 89 per cent use AI to generate or optimise written copy; 87 per cent report improved productivity; 58 per cent improved content quality; 39 per cent improved content performance. The productivity and performance figures are separate items with separate bases and are not presented in the essay as a single computed gap.

System1 and LinkedIn’s B2B Institute, research on B2B creative effectiveness, reported in Marketing Week, January 2021. 1,600 B2B ads tested with approximately six million people globally over four years. Findings used: 75 per cent scored one star or less; none reached five stars. Sample skews heavily to television advertising.

System1, eatbigfish and Peter Field, The Extraordinary Cost of Dull. Findings used: 60 per cent neutrality for UK B2B television advertising against 52 per cent for consumer; annualised share gain of approximately 1.3 points for non dull advertising against approximately 0.1 points for extremely dull at equivalent excess share of voice, reported as approximations and described in the essay as an order of magnitude rather than a precise multiple; of roughly £103 million in UK B2B advertising spend analysed, approximately £53 million sat in the extremely dull quartile.

Peter Field, The Crisis in Creative Effectiveness, IPA, 2019. Analysis of nearly 600 IPA case studies. Findings used: efficiency advantage of creatively awarded campaigns fell from approximately twelve times to below four times between 2008 and 2018; share of awarded campaigns that were short term rose from under 5 per cent to almost 40 per cent.

Anil R. Doshi and Oliver P. Hauser, “Generative AI enhances individual creativity but reduces the collective diversity of novel content”, Science Advances, July 2024. 293 writers randomised across three conditions, 600 independent evaluators. Findings used: novelty up 5.4 per cent with one AI idea and 8.1 per cent with five; usefulness up 3.7 and 9.0 per cent; collective similarity up 10.7 per cent of the measured range with one AI idea.

Joshua Ashkinaze, Julia Mendelsohn, Li Qiwei, Ceren Budak and Eric Gilbert (University of Michigan), How AI Ideas Affect the Creativity, Diversity, and Evolution of Human Ideas, ACM Collective Intelligence Conference, 2025. More than 800 participants across 40 countries. Finding used: high AI exposure increased collective idea diversity without affecting individual creativity.

Graphite, AI content volume research, May 2026. Method: 43,000 URLs randomly sampled from Common Crawl; detector validated at 4.2 per cent false positive and 0.6 per cent false negative rates. Observed share of new articles primarily AI generated was 49.9 per cent in Q1 2026 and 50.9 per cent in Q4 2025. Applying the published error rates puts the true share nearer 48 per cent, which is why the essay declines to treat the 50 per cent mark as a milestone and reports comparable volumes instead.

Graphite, AI content in search and answer engines, October 2025. Method: 31,493 keywords, first two pages of results. Findings used: 86 per cent of ranking articles human written; 7 per cent of top position results AI generated; approximately 82 per cent of AI answer engine citations human written. Correlation only; AI generated content may skew toward lower authority domains for unrelated reasons.

Ahrefs, AI Overviews click through research, May 2026, across 300,000 keywords. Finding used: 58 per cent average click through rate reduction for top ranking pages where an AI Overview appears, up from 34.5 per cent eight months earlier. Vendor research.

Pew Research Center, “Google users are less likely to click on links when an AI summary appears in the results”, July 2025. 68,879 searches tracked across a panel of 900 US adults, March 2025. Findings used: 8 per cent click rate with an AI summary present against 15 per cent without; 1 per cent click rate on links inside the summary. General consumer search rather than B2B specific.

SparkToro with Similarweb data, June 2026. Finding used: 68 per cent of US Google searches ended without a click, January to April 2026. General consumer search rather than B2B specific.

Nuremberg Institute for Market Decisions, Transparency without trust, Fabian Buder and Matthias Unfried, 2024. 3,000 respondents across the US, UK and Germany plus two controlled experiments. Findings used: identical advertising rated lower on emotional impact when labelled AI generated; 25 per cent believed they could recognise AI generated content, which is a confidence measure rather than a test of ability.

Gartner, press release May 2021, survey of more than 1,100 B2B customers conducted late 2020. Finding used: 64 per cent cannot distinguish one B2B brand’s digital experience from a competitor’s, with perceptions of difference between supplier offerings named as one of three drivers of buyer confidence.

Gartner, press release July 2019, survey of more than 1,000 B2B customers. Finding used: buyers experiencing information overload are 153 per cent more likely to settle for a smaller, less disruptive purchase than originally planned.

Jenni Romaniuk, Byron Sharp and Andrew Ehrenberg, “Evidence concerning the importance of perceived brand differentiation”, Australasian Marketing Journal, 2007. Finding used: low perceived differentiation across competing brands with purchase continuing regardless; distinctiveness proposed in place of differentiation. Not conducted in B2B categories, which are higher consideration than those studied.

Jenni Romaniuk, Ehrenberg-Bass Institute with LinkedIn’s B2B Institute, category entry points research. Concept used: buyers retrieve brands by buying situation, and candidate entry points are assessed on credibility, competitiveness and commonality.

6sense, B2B Buyer Experience Report for 2025, nearly 4,000 responses, and In It (Multiple Times) To Win It, November 2025. Findings used: 95 per cent of winning vendors were on the day one shortlist; around four in five deals went to the pre contact favourite; buyers reached first contact at 61 per cent of the journey, down from 69 per cent; 85 per cent of buyers had prior experience with the vendor selected. These describe winners rather than conversion odds.

John Dawes, Ehrenberg-Bass Institute with LinkedIn’s B2B Institute, The 95-5 Rule, 2021. Presented by the author as a heuristic rather than a measured constant.

Fabrizio Dell’Acqua, Edward McFowland III, Ethan Mollick, Hila Lifshitz-Assaf, Katherine C. Kellogg, Saran Rajendran, Lisa Krayer, François Candelon and Karim R. Lakhani, “Navigating the Jagged Technological Frontier”, Harvard Business School working paper 2023, published in Organization Science, 2025. 758 BCG consultants, AI access randomised, tasks classified as inside or outside AI’s capability frontier. Findings used: inside the frontier, 12.2 per cent more tasks completed, 25.1 per cent faster; outside it, 19 percentage points less likely to produce a correct solution.

Erik Brynjolfsson, Bharat Chandar and Ruyu Chen (Stanford Digital Economy Lab), Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence, August 2026 version. ADP payroll data covering millions of US workers. Findings used: employment of 22 to 25 year olds in AI exposed occupations 19 per cent below the counterfactual, revised up from 16 per cent in the November 2025 version; no comparable gap for experienced workers; declines concentrated where AI substitutes rather than complements. Covers AI exposed occupations across the economy rather than marketing specifically.

David S. Duncan, “How Do Workers Develop Good Judgment in the AI Era?”, Harvard Business Review, February 2026. Argument rather than original research.

Anu Atluru, “Taste is Eating Silicon Valley”, Working Theorys, September 2024. Argument rather than research. Definition used: taste is discernment expressed.

Digital marketing was one of the most important advances in the history of our profession, and I want to say that clearly before I say anything harder, because I have spent a large part of my career inside the machinery this essay describes and I write about it with as much affection as criticism.

Digital gave businesses direct access to their customers, made experimentation faster and cheaper, and lowered the cost of reaching a market. It let marketers observe behaviour rather than guess at it, personalise experiences that had once been generic, and connect activity more closely to commercial outcomes than the profession had ever managed before. Smaller organisations gained capabilities that had previously belonged only to companies with large media budgets, specialist agencies and established distribution, and the discipline became more accessible, more responsive and more accountable as a result.

It also had an unintended consequence, and the consequence is the subject of this essay. As digital channels became more measurable, the work of operating them became increasingly central to how marketing was organised, evaluated and understood, until the discipline gradually became associated with the platforms it used rather than the markets it served. Marketers became specialists in search, social, automation, analytics, content and conversion, while the broader questions of customer value, positioning, pricing, distribution and long-term demand quietly moved elsewhere in the organisation. Digital did not make marketing strategy less important. It made execution so visible, so immediate and so absorbing that many organisations lost sight of the larger discipline it was supposed to serve.

The result is a contradiction I have watched play out in rooms I was sitting in. Marketing has more data, more technology and more specialist capability than at any point in its history, yet many marketers have less influence over the fundamental decisions that determine whether their business grows. If I had to compress how it happened into a single sentence, it would be this one, and it will return through the essay: we optimised what we could see, and we lost sight of what we could not.

Understanding this history matters right now because we are at risk of repeating it with artificial intelligence. In our first edition Everyone Can Now Build. Few Know How to Market, I argued that AI is creating far more builders than marketers. In the last edition AI Isn't Replacing Marketers. It's Changing What Marketers Are Paid to Do, I went further: execution is becoming abundant, judgement is becoming scarce, and marketing’s value is shifting toward the people who can decide. This edition is the story of where that judgement went, because it did not disappear overnight and it was not taken from us. We traded it away through five quiet substitutions, each of which looked sensible at the time: the channel for the discipline, the measurable for the valuable, optimisation for diagnosis, capture for creation, and finally promotion for marketing itself. Once again a powerful new capability is making execution faster and more accessible, and once again the conversation is being dominated by tools, workflows and output, which is why it is worth asking what digital taught us about the difference between greater capability and better marketing before we accelerate any further.

Digital marketing was a genuine advance

Any critique of digital marketing should begin by acknowledging the scale of what it achieved, because the substitutions that follow were made by intelligent people responding to real progress.

The internet created new ways for organisations and customers to find one another. Search meant a business could appear at the exact moment a potential customer expressed a need. Websites became places where customers could research, compare and buy on their own terms. Email created a direct and inexpensive channel for ongoing communication, and social platforms opened forms of participation and audience-building that traditional media had never offered. Digital also transformed the economics of experimentation, because marketers could launch activity quickly, observe the response and adjust without waiting months for results, while smaller businesses could reach precisely defined audiences without buying national media.

Most importantly for what came next, digital gave marketing a stronger language of accountability. The profession had always struggled to demonstrate how its work contributed to business performance, and digital platforms provided an expanding set of measures that appeared to connect exposure, behaviour and commercial action in a single unbroken chain. Impressions could be counted, clicks could be tracked, leads could be attributed, conversion paths could be observed, and the cost of acquiring a customer could finally be compared with the revenue that customer generated.

This was real progress, and many of those measures remain valuable for understanding whether a campaign is functioning, finding friction in a customer journey, and improving decisions within a channel. The problem was never that marketing became measurable.

The problem was that the most measurable parts of marketing gradually came to represent marketing itself.

The first substitution: the channel for the discipline

The term “digital marketing” originally described a new set of channels and capabilities, and the distinction was useful while organisations were learning how websites, search, email and social advertising actually worked. Over time, however, digital developed its own professional identity, with new roles, teams, agencies, qualifications and bodies of expertise emerging around the operation of the channels themselves. Marketing departments were progressively divided into specialist functions covering paid search, paid social, content, lifecycle, automation, ecommerce, conversion optimisation and analytics.

Specialisation was necessary, and I want to be fair to it. These channels were complex, changed constantly and demanded technical depth: a search specialist needed to understand auction dynamics and keyword intent, an automation practitioner needed to understand data architecture and journey logic, and an analyst needed the skills to interpret enormous volumes of behavioural data. The difficulty arose when specialisation quietly fragmented the discipline it was meant to strengthen.

A customer does not experience a business as a collection of channel teams. They move through the product, the website, the sales conversation, the service experience, the advertising and the price as parts of a single relationship, and every interaction shapes their understanding of the company and their willingness to choose it. Inside the organisation, however, those same interactions sit in separate functions, each with its own objectives, systems and measures, so the search team optimises search, the social team optimises engagement, the website team improves conversion and the media team delivers its campaign targets. Every team can perform competently while the organisation as a whole has no clear answer to the questions that actually determine growth: which customers to prioritise, what value the business creates for them, what position it can credibly own, and how today’s demand capture should be balanced against the creation of tomorrow’s demand.

When marketing is organised around channels, the channel becomes the starting point for the decision, and the conversation begins with what should be done on the social platform, the search engine, email or the website rather than with the customer, the market and the commercial problem.

The capability begins to determine the strategy.

The second substitution: the measurable for the valuable

Digital did more than multiply the available measures; it changed the speed at which they arrived. Traditional marketing effects developed gradually and resisted isolation, while digital platforms offered feedback within hours, producing a constant stream of numbers that could be displayed on a dashboard and discussed at the next performance meeting. That visibility made digital activity far easier to manage, and it also handed short-term measures a decisive organisational advantage.

A click is observable, but a memory created by advertising is not. A lead can be entered into a system, while growing familiarity among buyers who will not purchase for years is almost impossible to record. A conversion can be connected to a campaign, while the cumulative effect of years of brand building, product experience, reputation and distribution cannot be assigned to any single source. When managers face pressure to demonstrate results, the visible measure becomes the defensible one, and investment drifts toward activity that produces an immediate signal even when that signal represents only a small part of how marketing creates value.

I have sat in the meeting where this drift happens, and it is worth describing because nothing about it feels like a mistake at the time. A modest brand investment is on the table beside a retargeting budget, and one of them arrives with a chart while the other arrives with a theory. The chart shows cost per conversion to two decimal places, the theory asks the room to believe in buyers nobody can name yet, and the budget follows the chart. Everyone behaves reasonably, nobody distorts anything, and the organisation still ends up somewhere nobody chose deliberately.

Les Binet and Peter Field documented exactly this tension in The Long and the Short of It, distinguishing short-term activation, which converts existing demand, from longer-term brand building, which creates future demand and broader commercial effects over time. The IPA’s summary of their research warns specifically that using very short-term online metrics as primary performance measures can damage long-term success. Their lesson is not that short-term metrics are meaningless, because a conversion rate can reveal whether a landing page works and cost per acquisition can help compare tactical options. The error occurs when these measures are promoted from indicators of specific activity into definitions of marketing effectiveness, because a campaign can achieve a strong click-through rate without changing customer behaviour in any commercially meaningful way, a channel can report an attractive return while taking credit for customers who would have purchased anyway, and a team can exceed its lead target while creating very little revenue or future demand.

The numbers may be accurate within the system that produced them, and the interpretation can still be wrong.

Metrics do not remain neutral

The second substitution deepens once measures stop being observations and start being targets, so it is worth staying with it a little longer. Once a metric becomes a target, people organise their work around improving it, an idea usually associated with Goodhart’s law: a measure that is useful as an indicator becomes less useful the moment it is turned into an instrument of control.

Marketing supplies endless examples of how this plays out in practice. When a team is rewarded for lead volume, it acquires an incentive to pursue audiences and tactics that generate inexpensive responses regardless of whether those leads will ever become valuable customers. When cost per acquisition dominates, investment concentrates on the people who are easiest to convert rather than those who represent the greatest opportunity. When return on advertising spend is treated as a complete measure of effectiveness, the channels that capture existing intent will always appear more productive than the ones that helped create the intent in the first place.

I have run a team that lived this pattern, and the memory still stings usefully. We beat our lead target for four consecutive quarters, celebrated each one, and watched the revenue number quietly travel in the other direction the whole time, because the system we had built was superb at generating the thing we counted and indifferent to the thing we needed. Nobody manipulated anything dishonestly; we simply made rational decisions inside the system we had been given, which is precisely how the distortion always works.

This is why measurement frameworks are strategic choices rather than administrative ones. They announce what the organisation values and they steer where resources flow, so a business whose every measure rewards immediate response will gradually assemble a marketing system designed to produce immediate response, one that becomes highly efficient at harvesting existing demand while growing progressively weaker at creating the conditions for future growth. Digital did not invent this problem, because businesses have always managed through imperfect proxies, but it multiplied the number and authority of those proxies while lending them an appearance of precision that exceeded what they could legitimately tell us.

Attribution offered certainty it could not always support

Attribution became one of digital marketing’s most compelling promises, because for the first time marketers appeared able to follow a customer from exposure to action and assign value to every touchpoint along the way, as though marketing investment could finally be evaluated with the precision of an accounting system. The difficulty is that observing an event before a purchase does not establish that it caused the purchase. A customer who searches for a company by name, clicks an advertisement and buys may have been shaped by years of prior experience, recommendations, brand advertising and product reputation, and the final click is visible only because it happened close to the transaction rather than because it created the demand.

The distinction between correlation and causation sits at the centre of marketing measurement, and the research here is sobering. A study published in Marketing Science by Brett Gordon and colleagues, using large-scale randomised experiments conducted on a major social platform, found that commonly used observational attribution methods often failed to recover the causal effects the experiments identified. Randall Lewis and Justin Rao reached an equally uncomfortable conclusion from 25 large field experiments with major US advertisers: the commercial effect of advertising is often so small relative to the natural variation in customer purchasing that even experiments involving millions of customers produced wide confidence intervals around return on investment. None of this means marketing cannot be measured, but it does mean the answer is rarely as simple as a platform dashboard suggests.

Les Binet has described how the early ability to count clicks encouraged systematic misattribution, because a click did not necessarily represent an incremental conversion, and his recommendation points the way forward: combine attribution, experiments and marketing mix modelling in a deliberately holistic approach, recognising that each method answers a different question and carries different limitations. Digital gave organisations vastly more information about observable behaviour, and the mistake was treating observable behaviour as a complete explanation of why customers acted. We optimised what we could see, and we lost sight of what we could not.

The third substitution: optimisation for diagnosis

Optimisation is one of digital marketing’s greatest strengths, because campaigns can be adjusted in response to performance, audiences refined, creative tested, friction removed and resources shifted toward whatever is working. But optimisation always begins with an existing system, improving the performance of a chosen activity against a chosen objective, and it cannot tell you whether the system itself was the right one to build. A team can lift the conversion rate of a landing page without questioning whether the proposition is strong enough, reduce the cost of a lead without examining whether the business is targeting the right market, and improve email engagement without asking whether customers find the communication valuable at all.

The distinction that matters is between optimisation and diagnosis. Diagnosis begins with the business problem, examining the market, the customer, the category, the product and the organisation before deciding which intervention is appropriate, whereas optimisation begins only after the intervention has already been chosen. Digital tilted organisations heavily toward the former because improvement could be observed continuously: the dashboard reported what had moved since yesterday, the platform offered its recommendations, and the next test was always ready to launch. Strategic diagnosis moves at a different pace, drawing on customer research, sales data, category behaviour, competitive analysis and commercial performance, and it carries an occupational hazard that optimisation never does, which is that it may reveal the problem does not belong to the marketing communications team at all. Perhaps the offer is undifferentiated, the price misaligned with perceived value, the product wrong for the intended segment, or the business simply difficult to buy from. Those findings cross functional boundaries and ask the organisation to reconsider decisions it believes it has already made.

It is always easier to adjust the campaign than to reopen a decision the organisation has already made.

The fourth substitution: capture for creation

Digital channels are exceptionally good at identifying and responding to existing intent. Search lets a business appear when a customer is actively looking, retargeting reaches people who have already visited, automation nurtures known prospects, and review platforms influence buyers who are actively comparing alternatives. These capabilities are commercially valuable, because a business should absolutely make it easy for interested customers to find, understand and choose it. The problem arises when demand capture is mistaken for the whole of marketing.

Most potential customers are not buying at any particular moment, especially in B2B categories with long replacement cycles. Professor John Dawes’ 95:5 rule uses a deliberately simplified ratio to make the point, and Dawes is careful to present 95 per cent as a heuristic rather than a law, but the implication holds across most categories: the large majority of your future buyers are out of market right now. They cannot be converted today no matter how precisely a platform can target them, because they already have a supplier, or lack an immediate need, or sit under contract without budget. What marketing can do is influence their future behaviour, because Dawes argues that advertising works principally by creating and refreshing brand-relevant memories that become useful when the buying situation eventually arrives, which is why a strategy focused only on buyers who are already searching will capture current demand while failing to build the mental availability that sustained growth requires.

Digital performance systems naturally favour the people closest to purchase, whose behaviour is easy to observe and whose response arrives quickly, so investment accumulates around search, retargeting, lead generation and conversion because those activities can demonstrate proximity to revenue. Yet proximity to revenue is not the same as responsibility for creating it. Capture and creation are complementary, and a business needs both; the danger is simply that one of them looks more accountable because its effects are easier to count.

The fifth substitution: promotion for marketing

The narrowing of marketing was never only a measurement problem, because over time it changed the function’s position within the organisation itself. Marketing in its broader sense concerns the creation, communication, delivery and exchange of value, which is the breadth the American Marketing Association’s definition still reflects, and Philip Kotler’s planning framework similarly begins with research and moves through segmentation, targeting, positioning and the value proposition before it ever reaches implementation. The discipline was designed as a sequence of choices about the market and the value an organisation intends to create, with execution arriving near the end.

In many modern organisations, marketing now enters near the end instead. Product teams determine what will be built, leadership selects the growth market, finance shapes the price, sales defines the priority accounts, and customer experience sits in a function of its own, after which marketing receives the finished offer and is asked to create awareness, generate leads and accelerate revenue. I know exactly how this feels from the inside, because I have been handed a product I first encountered at its launch briefing, carrying a price I was never consulted on, aimed at a segment chosen in a spreadsheet I never saw, and asked to make it famous. The team I led operated sophisticated technology and ran precisely targeted campaigns, and none of that capability bought us a seat at the decisions that actually determined whether the thing could grow.

This arrangement reduces marketing to promotion even when the department is technically excellent, and digital specialisation quietly reinforces it, because a function valued for operating the channels will be framed in terms of campaign delivery rather than market understanding. The organisation then meets a predictable problem: when growth falls short, marketing is asked to increase demand for an offer it did not help shape, among customers it did not select, at a price it did not set, and promotion becomes responsible for compensating for every weakness upstream. Stronger communication sometimes helps, but the constraint usually lives somewhere the campaign cannot reach.

What we lost sight of

The purpose of marketing did not change when customers moved online. Businesses still need to understand the markets they operate in, identify customer needs, decide which segments to serve and develop offers that create genuine value, and they still need to determine how those offers are positioned, priced, distributed and made easy to buy. They also still need to create demand beyond the customers who are ready to act today, which means building familiarity, trust and mental availability over time while maintaining the physical and digital availability to convert that demand when it finally emerges.

Digital capability can serve every part of that work. Search data can reveal how customers describe their own needs, website behaviour can expose friction, customer systems can illuminate buying journeys, online communities can surface category problems, and experiments can test propositions before serious money is committed. But these capabilities only become marketing when they are connected to a broader commercial purpose, and that connection is precisely what the five substitutions severed. A dashboard cannot determine the purpose, a platform cannot know which market the organisation should enter, and an attribution model cannot decide how much current revenue should be traded for future demand. Those decisions require marketers who understand the discipline beyond the operation of its tools.

Digital marketing is not the enemy

It would be easy to bend this argument into a rejection of digital marketing, and that would be both wrong and a little dishonest, because the problem was never digital. The problem is the separation of digital capability from marketing thought, and the responsibility for that separation is widely shared. Organisations rewarded immediacy, precision and visible activity, quarterly targets amplified the appeal of short-term response, structures fragmented responsibility for the customer, and technology companies promoted measurement systems that conveniently made their own platforms look accountable. Marketers played our part too, and I include myself in this: there were years when technical expertise felt like safer professional ground than difficult commercial judgement, because the expertise could be demonstrated on a dashboard while the judgement had to be defended in a room.

The industry also manufactured a false distinction between “traditional” and “digital” marketing, as though customer value, positioning, brand building and distribution belonged to an earlier era while data, automation and performance represented the future. Customers never experienced that distinction, because their behaviour always moved across physical and digital environments, and they encountered advertising, recommendations, salespeople, websites, products and service as parts of one relationship. The task was always to understand how those elements work together to create value and influence choice, which is why the answer is not a return to some pre-digital version of the discipline. The answer is to stop treating digital marketing as a substitute for marketing.

Reconnecting capability with cause

Reversing the five substitutions begins by placing the market and the customer ahead of the channel, which means that before deciding how to use search, social, content, automation or AI, a marketer needs a clear diagnosis of the growth problem and a considered view of whether the real constraint is awareness, relevance, distinctiveness, availability, customer experience, pricing, product fit or sales conversion. Measurement should then be designed around the commercial objective rather than inherited from the platform, which requires distinguishing between the metrics used to manage activity and the measures used to evaluate business impact, because clicks, engagement and cost per lead can improve execution without ever constituting evidence of incremental growth.

Different questions also demand different methods, because attribution supports continuous tactical optimisation, experiments establish whether an intervention actually changed behaviour, marketing mix modelling offers a broader view of how investment contributes over time, and customer and brand research detects effects that have not yet reached the sales data. No single method is sufficient, and better measurement comes from combining the evidence, understanding the limitations of each source and exercising judgement across them. Alongside better measurement, organisations need to restore marketing’s influence upstream, so that the customer and market perspective is represented in decisions about products, propositions, price and routes to market before the organisation commits to an offer and asks communications to rescue it.

Finally, marketing teams need broader foundations beneath their specialist expertise. Channel skills remain valuable, but they belong inside an understanding of how markets grow, how customers choose, how brands are built and how value is created, because the strongest search strategist understands demand beyond search, the strongest automation practitioner understands customers beyond the database, and the strongest performance marketer understands effects that no conversion window will ever capture. Specialists become more valuable, not less, when they understand the whole system their specialism operates within.

Before your next campaign plan

The argument above compresses into five checks, and they are worth running before the next brief is written:

1. Diagnose before you optimise. Name the actual growth constraint, whether awareness, relevance, distinctiveness, availability, price, product or conversion, before choosing the intervention, because a campaign without a diagnosis is a guess.

2. Design measurement around the objective, not the platform. Decide what business outcome the activity must change, then choose measures capable of detecting it.

3. Separate activity metrics from impact measures. Clicks and cost per lead help manage execution, and they should never be reported as though they were evidence of incremental growth.

4. Balance capture with creation, deliberately. Write down the split between converting today’s demand and building tomorrow’s, and defend it as a choice rather than inheriting it from last year’s dashboard.

5. Represent the market upstream. If the offer, the price or the segment is the real constraint, say so before promotion is asked to compensate for it.

The lesson for the AI era

The digital era taught marketers to associate progress with greater speed, more data and tighter optimisation, and AI is now amplifying every one of those capabilities at once. It can produce more content, analyse larger datasets, automate more decisions and accelerate execution to the point where a small team can operate with the output of a large one, which is an enormous opportunity and also a sharpened risk, because AI raises the cost of acting without a clear marketing foundation. A weak proposition can now be promoted at unprecedented scale, an undifferentiated position can generate hundreds of content variations, a poorly chosen metric can be optimised continuously, and a fragmented customer experience can be automated end to end.

Left undirected, AI will intensify exactly the habits that narrowed marketing through the digital era, making organisations ever more efficient at producing and measuring activity while remaining unclear about whether any of it contributes to growth. The lesson is not to slow the technology down but to strengthen the judgement directing it, because digital gave us remarkable capabilities and we forgot that capability is not strategy, that activity is not effectiveness, and that a measurable response is not the same thing as customer value. AI now offers us the chance to learn from that mistake while it still costs little to do so.

Remembering marketing

Marketing has never been defined by the tools available at any particular moment. Its purpose is to understand markets and customers, create value, build demand and help the organisation make better commercial decisions, and channels, platforms and technologies matter precisely because they serve that work. In the last edition I described marketing as the orchestration of growth, and the digital era, for all it gave us, was the period when much of the profession put down the conductor’s baton and picked up a single instrument. We played it with real skill, and the orchestra gradually stopped expecting us to conduct.

Digital marketing should have expanded the influence of the discipline, and in many organisations it narrowed it instead, leaving the profession more technically capable while drawing it steadily downstream, further from the choices that shape the business. Recovering the broader discipline does not require abandoning digital expertise; it requires placing that expertise back inside marketing, where the starting point is the market rather than the platform, the objective is commercial progress rather than activity, and the role of measurement is to sharpen judgement rather than to replace it. The role of marketing, in turn, was never simply to promote whatever the business decided to build, but to help the business decide what value it can create, for whom, and how that value becomes a source of growth that lasts.

Digital did not make us forget marketing on its own; we allowed the tools, the metrics and the pace of execution to consume attention that belonged to customers, markets and strategy, and we optimised what we could see until we lost sight of what we could not. The next era hands us the same choice at a faster tempo. We can use AI to produce more of the same at greater speed, or we can use it to lift the execution load off the profession and return our attention to the work that was waiting underneath it the whole time. That is not a rejection of digital marketing. It is the completion of its promise.

Evidence informs. Judgement decides.

This is the version of marketing FP Collectiv teaches. Not channel tricks and tool tours: the evidence, frameworks and commercial judgement to diagnose growth problems, design measurement that answers real questions, and defend the strategy in the room that matters. AI is taught as a layer on top of those fundamentals, never a substitute for them. Explore the courses and membership → FP Collectiv

Sources

• American Marketing Association, Definitions of Marketing, on marketing as the creation, communication, delivery and exchange of value.

• Philip Kotler, The Past, Present, and Future of Marketing, on marketing research, segmentation, targeting, positioning and value creation as part of the planning process.

• Les Binet and Peter Field, The Long and the Short of It, IPA, on the distinction between short-term response and long-term brand building, and the risks of relying on short-term online measures.

• Les Binet, Unlocking Marketing Effectiveness in the Digital Age, on misattribution, incrementality and the complementary roles of attribution, experiments and marketing mix modelling. (Verify exact publication title and outlet before publishing.)

• Brett Gordon, Florian Zettelmeyer, Neha Bhargava and Dan Chapsky, A Comparison of Approaches to Advertising Measurement, Marketing Science, on differences between observational measurement and randomised advertising experiments.

• Randall Lewis and Justin Rao, The Unfavorable Economics of Measuring the Returns to Advertising, on the statistical difficulty of estimating advertising returns.

• John Dawes and the Ehrenberg-Bass Institute for Marketing Science, Advertising Effectiveness and the 95:5 Rule, on out-of-market buyers and advertising’s role in building future brand-relevant memories.

• Charles Goodhart, Problems of Monetary Management: The UK Experience, on the instability created when observ

The question I am asked most often about artificial intelligence is whether it will replace marketers. I think that is the wrong question.

Not because the worry behind it is misplaced. It isn’t. I have spent my career inside marketing teams, and I have never seen a technology land the way this one has. Not because it arrived suddenly, but because of where it arrived. Every previous wave of marketing technology gave us new places to do the work or new ways to measure it. This one does the work. That is a different kind of arrival, and it is why so many capable people are quietly looking at their own role and wondering how much of it a machine could now do.

The question is wrong because it asks for a yes or a no, and the truth is not a yes or a no. It is a movement. Here is the claim this essay will try to earn: the centre of gravity in marketing is shifting, away from executing the work and toward orchestrating the growth the work exists to produce. And here is why, in eight words:

Execution is becoming abundant. Judgement is becoming scarce.

Hold that claim against the strangest finding I have seen in this year’s research. Gartner’s 2026 CMO Spend Survey, which polled 401 marketing leaders at the start of this year, found that nearly two thirds of marketers believe AI will transform their roles. Only 32 per cent believe they need to update their skills.

Sit with that gap for a moment, because it is doing a lot of work. Two thirds of a profession can see the wave. Two thirds of the same profession has concluded it applies to someone else. I don’t think that is arrogance. I think it is a category error. Most marketers are asking “will I be replaced?”, getting the correct answer, no, and mistaking it for reassurance. Very few marketers will be replaced outright. A great many will find that the part of the job that used to define their value is quietly becoming the part a machine does best. The value does not disappear. It moves.

To see where it moves to, and what that means for your career and your team, you have to see this moment for what it is: not a new tool, but the fourth structural shift in how marketing creates value. The first three, digital, attribution and the technology stack, built the profession you work in today. AI is the fourth, and it behaves differently from all of them. It is worth walking through the sequence, because the pattern the first three reveal tells you almost everything about what happens next.

Four Shifts, One Direction, Until Now

For most of its history, marketing was a generalist craft. Before digital, one person, or one small team, understood the customer, shaped the message, chose where it ran, and judged whether it worked. The tools were few and stable: print, broadcast, direct mail, the trade show, the sales force. What made a marketer valuable was not mastery of any surface, because the surfaces barely changed. It was judgement about customers and messages, exercised with patchy data and a long feedback loop. You ran the campaign, you waited, you argued in a meeting about whether it worked. The craft was slow, but it was whole. One mind held the entire system.

The first shift broke that wholeness: digital multiplied the surfaces. Search, then social, then mobile, then video, each arriving with its own mechanics, its own auction dynamics, its own metrics, its own learning curve. No one mind could hold it all any more, and organisations responded rationally. They specialised. The generalist gave way to the paid search specialist, the SEO lead, the automation manager, the social team. This was not a mistake or a fashion. It was the correct response to genuine complexity. When running a single channel well takes real expertise, and doing it badly wastes real money, you hire someone who does only that.

The second shift was attribution, and it was subtler, because it changed not what marketers did but what they were believed for. When digital made activity trackable, the monthly argument about whether the campaign worked became a dashboard. That bought marketing credibility it had never had, and I would not give it back. But it came with a cost we are still paying: what could be measured became what mattered. The channels that could prove their contribution attracted the budget, the headcount and the ambitious people. The work that compounds slowly and resists measurement, brand, memory, reputation, was quietly starved. Attribution did not just measure the work. It reshaped the work in its own image.

The third shift was the technology stack itself. When Scott Brinker published his first marketing technology landscape in 2011, it held roughly 150 tools. By 2024 it counted more than 14,000. A hundredfold increase in barely more than a decade, and every one of those tools needed someone to own it, integrate it, and justify its licence fee. Operating the machinery became a career in itself.

Step back and look at what those three shifts have in common. Each one moved marketing’s centre of gravity closer to execution. Each one made the profession more specialised, more measurable, more operational. Headcount, budgets, career ladders and status were all progressively reorganised around the ability to produce and run the work. Being good at marketing came to mean being good at executing a discipline within it. And to be clear, that structure served us well. The people who built their careers this way were responding correctly to the world as it was.

The fourth shift is AI, and here is the thing I most want you to see: it runs in the opposite direction. For thirty years, every structural change pushed value toward execution. AI is the first that pushes it back. The trouble is that today’s marketing team was designed for yesterday’s constraints. It was built for a world where execution was expensive, slow and scarce. AI changes exactly those constraints, and a structure built on constraints that no longer exist does not stay standing out of respect for its history.

The Economics Underneath the Noise

Set aside, for a moment, the demos and the discourse. The strategically important point is simpler and more durable than any feature list: AI collapses the cost and time of a large share of marketing execution.

This is no longer a claim you have to take on faith. In a 2023 experiment published in Science, MIT researchers gave 453 professionals realistic business writing tasks; those with access to an AI assistant finished 40 per cent faster and produced work that blind assessors rated 18 per cent higher. Harvard researchers working with Boston Consulting Group ran a larger field experiment: 758 consultants, realistic strategy and marketing tasks, and the group using a frontier AI model completed more tasks, faster, at quality rated 40 per cent higher than the control group.

I believe those numbers, because I have watched them happen. Work I once scoped in fortnights comes back in afternoons. The first draft, the variant set, the campaign summary, the first cut of a segment: the honest response of anyone who has run a team through the last two years is not scepticism about the productivity data. It is recognition.

But the number in the Harvard study that should actually change how you plan your career is not the speed figure. It is this: the biggest gains went to the bottom half of performers, whose output improved by 43 per cent, far more than the top performers gained. Think about what that means. AI is not a tool that makes the best people better, mostly. It is a tool that makes average execution nearly indistinguishable from good execution. The gap that justified the premium for competent production, the gap between the adequate specialist and the strong one, is compressing in front of us.

And when average output rises everywhere at once, something counterintuitive happens: the market stops paying for it. This is where basic economics tells you what no feature demo will. When a resource becomes abundant, its relative value falls. Not because it stops mattering. Water matters more than diamonds, and costs less, because scarcity is what commands a premium. For thirty years, the scarce resource in marketing was the ability to produce and run the work. That is the resource AI just made abundant.

Execution is becoming abundant. Judgement is becoming scarce.

Follow that sentence one step further and you get the whole restructuring. When production was the bottleneck, you needed many people spending most of their time producing. As production stops being the bottleneck, the constraint moves to a different question entirely: not who can make the work, but who can decide what work is worth making. Organisations will need fewer hours spent producing, and more spent deciding. Everything else in this essay follows from that.

Where the Value Is Pooling

If the economics are right, we should already see the redistribution in the labour market. We do, and it is worth being honest about what it looks like, because “AI is not replacing marketers” should be a conclusion drawn from evidence, not a comfort blanket.

The most careful employment study so far comes from Stanford’s Digital Economy Lab, where Erik Brynjolfsson and colleagues analysed payroll records covering millions of American workers. Their finding is precise and uncomfortable: since generative AI became widespread, early-career workers aged 22 to 25 in the most AI-exposed occupations, marketing squarely among them, have seen a roughly 16 per cent relative decline in employment. But the same study found that employment for experienced workers in those same occupations held stable or grew.

Read those two findings together, because together they are a map of where the value is going. The damage is not spread evenly across the profession. It is concentrated exactly where the job consists of executing well-defined tasks, which is what entry-level roles were built around, and which is what the models now do cheaply. Where the job consists of judgement built on experience, employment is holding. The labour market is not saying “marketers are finished.” It is saying, with unusual clarity, “we will keep paying for judgement, and we will stop paying for undifferentiated execution.” Indeed’s Hiring Lab, analysing the skills listed in millions of job postings, reached the same destination from a different direction: nearly half the skills in a typical posting can be transformed or heavily assisted by generative AI, fewer than one per cent can be fully performed by it, and the roles that remain shift, in their words, from doing the work to directing it.

This is what I mean when I say the centre of gravity is shifting. For a decade, a marketer’s value sat close to the work itself: producing assets, running campaigns, optimising channels. That is the layer AI is absorbing fastest. As it does, value moves to the layer above: understanding customers, defining strategy, allocating resources, integrating across functions, and making the commercial calls that decide whether any of the activity was worth doing.

Be careful how you read this, though, because it is not an argument against expertise, and it is emphatically not “specialists out, generalists in.” A world where execution is cheap is a world with more execution in it, not less, which means more need for people who deeply understand how a discipline behaves, where it breaks, and what good looks like when a machine produces it at volume. The Harvard experiment carries a warning here that I think about often. The researchers included a task that looked like something AI would handle well but sat just beyond its real capability. Consultants without AI got it right 84 per cent of the time. Consultants with AI got it right 60 to 70 per cent of the time. The tool made them faster, more confident, and wrong more often, because they stopped interrogating the output. The researchers called it falling asleep at the wheel. Cheap execution without expert judgement does not merely fail to add value. It subtracts value, at scale, with total confidence.

So the rising value is not in generalism. It is in the connective tissue. When every channel can be run faster and cheaper, the hard part is no longer running any single one of them. The hard part is deciding how they fit together: how paid, owned and earned combine, how brand and performance trade off, how this quarter’s number is weighed against next year’s growth. The scarce skill is connecting expertise across disciplines into one coherent commercial system, rather than a stack of well-run silos pointing in slightly different directions.

That is a different job from the one most marketing ladders were built to reward. And it has a name.

The Growth Orchestrator

Here is the shift stated plainly. Marketing is the orchestration of growth. It does not create growth in isolation, and it never did. It orchestrates a system that creates growth.

I want to earn that sentence rather than assert it, so picture a team I suspect you will recognise, because I have sat inside versions of it more than once. The paid media is well run. The content calendar ships on time. The automation flows fire. The SEO trend line points the right way. Every specialist is hitting their number, every dashboard is green, and the company still misses its revenue target. Everyone has done their job, and somehow the sum of the jobs does not add up to growth. When the post-mortem comes, no one in the room can quite explain why, because the explanation does not live in any single channel. It lives between them. Demand was captured efficiently but never created. The story sales told was subtly different from the story the campaigns told. The budget followed last quarter’s attribution instead of next year’s buyers. No individual was wrong. The system was.

That gap, between well-executed parts and a working whole, is where marketing’s real job has always lived, even when our org charts pretended otherwise. The growth system has six stages, and it is worth naming them now, because the essays ahead in this series will take each one in turn. It begins with market and customer understanding: knowing who buys, why, and what they are really weighing when they decide. That feeds strategy and positioning: choosing where to compete and what you intend to be remembered for. Then comes demand creation, building memory and preference among the buyers who are not yet in the market, and demand capture, converting the few who are. Then conversion support, arming sales so the promise survives contact with the buying committee. Then retention and expansion, keeping and growing the customers you fought to win. And the loop closes with learning, feeding what the market just taught you back into the understanding you started with. Marketing’s job is not any single stage. It is the coherence of all of them.

For thirty years we could not staff that job properly, because execution consumed everything. The channels had to be fed. The stack had to be run. The reporting had to be built. Coordination was the thing you did in the gaps, in the weekly meeting, in the planning cycle, if there was time, and there was never time. The profession’s deepest work was perpetually crowded out by its busiest work.

This is the possibility that AI actually opens, and it is why I refuse to read this moment as a diminishment. AI does not sit outside the growth system. It sits inside it, and it can now support nearly every stage: sharpening the research, drafting the positioning, producing the demand-creation work, optimising the capture, arming the sales team, personalising the experience. What it cannot do, structurally cannot do, is decide where to compete, how to position, what to prioritise, and how to weigh this quarter against the long term. It runs inside the system. Someone still has to conduct.

That someone is the growth orchestrator: the person who holds the whole system in view, decides where effort and money should go, and makes the trade-offs no single specialist and no model is positioned to make. Not the channel manager with a bigger title. A genuinely different role, closer to what a conductor does with an orchestra of virtuosos: not playing any instrument, but responsible for the only thing the audience actually hears, which is whether it all comes together.

And the timing is not incidental, because the buyer’s side of the system has become dramatically harder to serve with silos. Forrester’s 2026 buying research found the average B2B purchase now involves around 13 internal stakeholders and 9 external influencers, and that buyers increasingly begin their research inside AI tools, then validate what the machines tell them with sources they already trust: peers, communities, and the brands that had earned a place in their memory long before the buying process began. Your buyer is running an AI-assisted evaluation across a committee of twenty. A marketing function organised as independent channel silos cannot meet a buying process that behaves as a system. Only a system meets a system, and someone has to run yours.

Orchestration is also, now, a resource-allocation job in the most literal sense. Gartner reports AI already claims 15.3 per cent of the average marketing budget, while 56 per cent of CMOs say they lack the budget to execute their strategy at all. Scarce money plus abundant execution is precisely the condition under which allocation, not production, decides who grows. Someone has to choose what the AI spend displaces, which capabilities it builds, and how its contribution will be judged. Those are orchestration decisions, and in most organisations today, nobody fully owns them: the same Gartner survey found 70 per cent of CMOs naming AI leadership as a key goal while only 30 per cent say their organisation is ready to scale it. That gap is rarely a technology problem. It is a conducting vacancy.

Here is what I find energising rather than threatening about all of this. The orchestrator role is not a consolation prize for people whose real job got automated. It is the job the best marketers always wanted and were rarely allowed to do, the one buried under the execution load: to actually understand a market, actually shape where the business plays, actually decide. If you became a marketer because you wanted to grow something, rather than because you wanted to operate software, this shift is not taking your job away. It is finally offering it to you.

That system is the organising lens for the rest of this series, and it deserves more than a few paragraphs. We will give it an essay of its own and walk each stage in full. For now, hold the shape: a connected system that produces growth, that AI increasingly runs within, and that a human increasingly has to conduct.

One long-form edition at a time, on making better marketing decisions. Free to subscribe.

Subscribe now

Judgement, and Taste

If evidence is more accessible than ever, analysis faster than ever, and execution cheaper than ever, then none of those is where your edge lives. What is left, and what does not automate cleanly, is the ability to decide between competing priorities when there is no objectively correct answer.

Take the most familiar trade-off in demand generation: how much to invest in creating future demand versus capturing the demand that exists today. The evidence here is about as good as marketing evidence gets. The Ehrenberg-Bass Institute’s work on mental availability shows why being remembered by buyers who are not yet in the market matters. John Dawes’s research with the LinkedIn B2B Institute, usually shortened to the 95:5 rule, gives the arithmetic: at any moment, roughly 95 per cent of business buyers are not in the market for your category, so most of your marketing works, if it works at all, by building memory for a purchase that has not started yet. Les Binet and Peter Field, analysing hundreds of campaigns in the IPA’s databank, showed the long-term cost of over-investing in short-term activation, and proposed a rough 60:40 balance of brand to activation, with their B2B work pointing nearer an even split.

Now notice what all of that evidence cannot do. It tells you what tends to work, on average, across categories. It cannot tell you how much to weight brand against activation in your market, for your category, against your targets, this year, and I can tell you from experience that the moment this question stops being academic is not comfortable. It arrives in a room, near the end of a hard quarter, when someone senior points at the brand line of the budget and asks what it did for pipeline this month. Every marketer who has held a budget knows that moment. The research is on your side, and the research is not enough, because the honest answer involves your cash position, your board’s patience, your competitor’s noise and your own conviction. An AI assistant can recite Binet and Field flawlessly. It cannot tell you whether your business can survive the eighteen months brand investment takes to pay back, because that answer is not in the training data. It is in you, or it is nowhere. Evidence informs. Judgement decides.

There is a second capability beside judgement, and we are about to need much more of it. Call it taste. Judgement determines what should be done. Taste determines how well it must be expressed, and whether this particular expression clears the bar. In a world where anyone can generate a competent asset in seconds, competent stops being a differentiator; the research says the floor is rising everywhere, for everyone, including your competitors. When the floor rises, the only advantage left is the distance between the floor and your ceiling. I have reviewed plenty of machine-drafted work this year that was fluent, on-brief, and entirely forgettable, fine, but not us, and the discipline of saying “not yet” to work that is merely fine is becoming one of the last durable advantages a marketer has. Taste is also the guard against falling asleep at the wheel: someone in the system has to be the person who looks at confident, plausible output and catches the one that is confidently wrong. In an AI-enabled team, that person is not a luxury. That person is the quality system.

What to Build If You Want to Stay Valuable

The obvious response to all this is to go and learn the tools. It is also the wrong first move, or at least an incomplete one. Tools change every few months, and fluency with this quarter’s features is the fastest-depreciating asset in your stack. It is telling that only 32 per cent of marketers believe they need to update their skills: the temptation for the minority who do act will be to close the gap with prompt tricks and tool tours, because those are easy to buy and easy to tick off. But the durable investment is to deepen the capabilities that make your judgement worth having in the first place. Five stand out.

First, customer and market understanding deeper than a model can infer from a prompt. AI can summarise the research you give it. It cannot sit in a sales call and hear the objection under the objection. Marketers who keep doing primary work, talking to customers, listening to calls, reading the market first-hand, hold an advantage over everyone whose understanding comes from prompting the same models as their competitors.

Second, commercial and financial literacy. Understand the business you sit inside: its economics, its margins, what a customer is worth and costs to acquire. The marketers who allocate resources and defend budgets are the ones who speak in commercial terms, and that conversation is getting harder as boards start asking what the AI spend is returning.

Third, the evidence base of marketing itself. Ehrenberg-Bass on how brands grow. Binet and Field on effectiveness. Behavioural science on how buyers actually decide, as opposed to how our funnel diagrams say they should. This knowledge compounds, because it is the raw material judgement is made from, and it is what lets you evaluate machine output rather than merely accept it.

Fourth, orchestration itself: the unglamorous, deeply human work of getting sales, product, finance and marketing pointed at the same growth thesis. Thirteen insiders and nine outside voices in the average buying group is also a fair picture of your internal reality. Systems are run by coalitions, and building coalitions is a skill no model has.

Fifth, and only now, AI fluency. It belongs on the list; working knowledge of what the tools do well and where the frontier currently runs is becoming table stakes. The point is not to skip it. The point is to sequence it. AI fluency multiplies the value of judgement, market knowledge and commercial skill, and multiplies nothing when they are absent.

This is also the honest part, so let me be honest. This shift is not painless, and it is not automatic. Marketers whose identity is built on executional craft are the most exposed, and the employment data says the squeeze is already real at the entry level, which means leaders have a second obligation the essay above implies but I want to make explicit: the old bottom rung of the ladder is gone, and we will have to build our juniors a new one, deliberately, or the next generation of judgement never gets formed. Moving up into orchestration is a real change in skill, not a title bump. Not everyone will make the move, and pretending otherwise helps no one. But the direction of travel is clear enough to act on now, while it is still a choice rather than a scramble.

The Work, Revealed

So, will AI replace marketers? It was the wrong question all along, and by now you can see why. AI will keep removing tasks, and some of those tasks have defined entire jobs. That is real, and it deserves respect rather than a wave of the hand.

But a task is not the work. Marketing has always been about creating value for customers and growth for the organisations they buy from. Every structural shift before this one, digital, attribution, the stack, buried that purpose a little deeper under the mechanics of doing. AI is the first shift that digs it back out. Strip away the drafting, the assembling, the reporting, the operating, and what remains is what was underneath the whole time: understanding markets, making sound decisions under uncertainty, and bringing people, ideas and capabilities together in pursuit of growth that lasts. The machine has not automated the work. It has revealed it.

So if you are carrying the worry, and most of us are, let me offer you something sturdier than reassurance. The worry comes from looking down at your tasks and watching some of them walk away. The hope comes from looking up at the work, and remembering that the tasks were never why you chose this. You did not become a marketer to operate software. You became one to understand people, to make calls that matter, to grow something. For thirty years the profession asked you to do that in the gaps between the busywork. The busywork is leaving. What remains is the job you actually wanted.

Organisations rarely remember who executed the work. They remember who changed the trajectory of the business.

Execution is becoming abundant. Judgement is becoming scarce. The centre of gravity is moving toward the people who can decide. Make sure you are standing where it lands.

This is the second piece in a series on making better marketing decisions in an AI-enabled world. If it was useful, follow along: the essays ahead go deeper on the growth system underneath all of this, and on how to build the judgement the next decade will pay for.

New here? Subscribe for the next edition.

Subscribe now

Sources

• Gartner, 2026 CMO Spend Survey, published May 2026. 401 CMOs and marketing leaders surveyed January to March 2026; AI at 15.3 per cent of marketing budgets; 70 per cent name AI leadership a key goal; 30 per cent ready to scale AI capabilities; 56 per cent report insufficient budget; roughly two thirds of marketers expect AI to transform their roles while 32 per cent believe they need to update their skills.

• Shakked Noy and Whitney Zhang, “Experimental Evidence on the Productivity Effects of Generative Artificial Intelligence”, Science, 2023. 453 professionals; 40 per cent faster completion, 18 per cent higher rated quality on professional writing tasks.

• Fabrizio Dell’Acqua et al., “Navigating the Jagged Technological Frontier”, Harvard Business School working paper with Boston Consulting Group, 2023 (published in Organization Science, 2025). 758 consultants; faster completion and quality rated 40 per cent higher inside the frontier; bottom-half performers improved 43 per cent; accuracy fell from 84 per cent to 60-70 per cent on the outside-frontier task.

• Erik Brynjolfsson, Bharat Chandar and Ruyu Chen, “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence”, Stanford Digital Economy Lab, 2025. ADP payroll data; ~16 per cent relative employment decline for workers aged 22 to 25 in the most AI-exposed occupations; stable or growing employment for experienced workers in the same occupations.

• Indeed Hiring Lab, “AI at Work Report 2025: How GenAI is Rewiring the DNA of Jobs”, September 2025. ~46 per cent of skills in a typical US job posting subject to hybrid or full transformation; fewer than 1 per cent fully performable by GenAI today; workers shifting from doing the work to directing the work.

• Forrester, “The State of Business Buying, 2026”, January 2026. Average buying group of ~13 internal stakeholders and ~9 external influencers; AI tools increasingly the starting point of buyer research; buyers validating AI-generated answers with trusted human sources.

• John Dawes, Ehrenberg-Bass Institute for Marketing Science with the LinkedIn B2B Institute, “The 95:5 Rule”, 2021. At any given time, up to 95 per cent of business buyers are not in the market for a given category.

• Les Binet and Peter Field, “The Long and the Short of It”, IPA, 2013 (IPA effectiveness databank); and Binet and Field with the LinkedIn B2B Institute, “The 5 Principles of Growth in B2B Marketing”, 2019. ~60:40 brand-to-activation balance overall; closer to an even split indicated for B2B, labelled tentative by the authors.

• Scott Brinker, chiefmartec, Marketing Technology Landscape Supergraphics, 2011-2025. ~150 tools in 2011; 14,106 in 2024; 15,000+ in 2025.

Every generation of marketers believes it is living through a transformation of the profession. Television changed how brands reached mass audiences. Databases introduced new forms of targeting and direct response. The internet changed distribution, search changed discovery, social media changed participation, and smartphones placed all of it in our hands.

Now it is artificial intelligence.

But this is not another article about AI. It is an article about what becomes more important when the ability to create is no longer the main constraint.

The current wave deserves much of the attention it is receiving. Generative AI is not simply another media channel or software upgrade. It is changing who can create, how quickly ideas can be developed and how much work a small team, or even one person, can accomplish.

A founder can describe an application in plain language and produce a working prototype without first assembling a development team. A subject-matter expert can build a website, create a course, publish a newsletter and produce credible marketing materials using tools that would have been inaccessible or prohibitively expensive only a few years ago. Marketers can use AI to support research, analysis, writing, coding, design and administration.

The productivity evidence is already meaningful. In an experiment involving professional writing tasks, researchers Shakked Noy and Whitney Zhang found that people using generative AI completed their work 40 per cent faster, while the quality of their output improved by 18 per cent. A separate field experiment involving 758 Boston Consulting Group consultants found that AI users completed more tasks, worked faster and produced higher-quality results when the work fell within the technology’s capabilities.

These are significant gains. They suggest that capabilities once confined to specialists or larger organisations are becoming available to many more people.

And that is exactly why the question that matters has changed.

The barrier to building is falling

For much of modern business history, turning an idea into a product required considerable resources. Software needed engineers. A credible brand needed designers and agencies. Manufacturing required capital and access to suppliers. Reaching customers at scale required distribution relationships or substantial media budgets.

Those barriers have not disappeared, and there remains an enormous difference between producing a prototype and building a secure, differentiated and commercially sustainable product. Expertise still matters. Quality still matters. Capital still matters.

But the starting point has changed.

More people can now move from an idea to something tangible. The time and cost required to test a concept are falling, and work that once demanded a team can increasingly be attempted by an individual. In one controlled GitHub experiment, developers using its AI coding assistant completed a programming task 55 per cent faster than those working without it. The technology did not remove the need for engineering expertise, testing or judgement, but it reduced some of the effort involved in translating an instruction into functioning code.

As the cost of creating something falls, more things will be created. Markets will fill with more applications, services, courses, newsletters, consultancies, products and content. Many will be competently made. Some will be excellent.

The supply of attention, however, does not increase simply because the supply of things competing for it has multiplied.

Lower barriers to production therefore do not remove the commercial challenge. They change its location. When creating something becomes easier, the greater challenge is knowing what to build, who to build it for and why it will matter to them. It is understanding how to make an unfamiliar product relevant, how to earn attention in a crowded market and how to turn a first visitor into an audience, an audience into customers and a promising product into a sustainable business.

These are not primarily technology questions. They are marketing questions.

The bottleneck has moved

Much of the conversation about AI ends at the moment of creation. Someone builds an application over a weekend, produces a website in an afternoon or generates a month of content in minutes. The speed is impressive, but the harder commercial questions remain unanswered.

A product can be technically possible without solving an important problem. A website can look credible without giving anyone a reason to visit. A business can publish frequently without developing an audience. An automated email sequence can operate perfectly while promoting an offer nobody particularly wants.

The technology can reduce the effort required to bring an idea into the world. It cannot establish that a sufficient number of people experience the problem, are dissatisfied with the alternatives, are willing to change their behaviour and can be reached economically. Nor can it determine whether the offer is sufficiently distinctive, whether buyers will trust an unfamiliar provider or whether the economics will support growth.

Theodore Levitt warned against this tendency decades ago in Marketing Myopia. His argument was that businesses become vulnerable when they define themselves by the products they make rather than by the customer needs they serve. The danger is even greater when a product can be built before the market has been properly understood.

The ability to build is not evidence of demand.

A founder may be able to launch globally from the first day, but global availability does not produce global awareness. A product may be capable of serving a million customers, but scalability is irrelevant if the first hundred cannot be acquired. The fact that something exists does not mean that people will notice it, understand it, remember it, trust it or choose it.

The bottleneck has moved from producing the product to creating the market conditions in which the product can succeed.

One long-form edition at a time, on making better marketing decisions. Free to subscribe.

Subscribe now

Marketing was never just promotion

One reason this shift is misunderstood is that marketing has gradually been reduced to a narrow set of visible activities.

In many organisations, marketing has become shorthand for advertising, social media, content, email, campaigns and lead generation. These are legitimate parts of marketing, but they are not the whole discipline.

The American Marketing Association defines marketing as the activities and processes involved in “creating, communicating, delivering, and exchanging” offerings that have value. Communication is part of that definition, but it sits alongside the creation and delivery of value.

The traditional Four Ps make the same point. Product, Price, Place and Promotion describe a much broader commercial remit than communications alone. Promotion is one component of the marketing mix, not the definition of marketing.

Philip Kotler’s work helped establish marketing as the process of understanding markets, identifying customer needs, creating value and choosing which customers an organisation is best equipped to serve. Marketing, in this conception, begins well before a campaign is briefed. It influences the product, the market, the customer proposition, the route to market and the commercial model.

This places marketing upstream in the business. It should help determine which problems are worth solving, how a market should be understood, which customers represent the best opportunity and why an organisation has a credible right to win. It should contribute to decisions about the offer itself, the price customers will accept, the way the product is distributed and the position the business can occupy in the customer’s mind.

Promotion comes after those decisions. It cannot compensate for their absence.

Yet many marketers are brought into the process only after the product has been developed, the commercial target has been set and the launch date has been agreed. Product teams build, finance sets the price, sales defines the target audience, and marketing is asked to generate demand.

When marketing is confined to the final stage, its role is no longer to help the organisation understand and shape the market. It is to communicate decisions that have already been made.

AI did not cause this narrowing of the discipline. It began much earlier.

Digital marketing played an important part in accelerating it.

How digital marketing changed what we valued

Digital marketing brought enormous benefits to the profession. It expanded access to customers, made experimentation faster and allowed smaller businesses to reach markets that had previously been inaccessible. It created new forms of customer experience and gave marketers far greater visibility into behaviour and performance.

It also made marketing more accountable. For organisations that had struggled to connect marketing investment with commercial outcomes, this was an important advance.

But every new capability influences what people pay attention to. Digital platforms provided an immediate stream of measurable activity: impressions, clicks, open rates, conversions, leads, cost per acquisition and return on advertising spend. Because these metrics were visible and available quickly, they became easier to report, optimise and defend.

The problem was not measurement. The problem was allowing what could be measured immediately to define the value of marketing.

Les Binet and Peter Field identified this tension in their analysis of the IPA Effectiveness Databank. Their research distinguished between short-term activation, which helps convert existing demand, and long-term brand building, which creates future demand and produces broader commercial effects over time. They also warned that the growing reliance on very short-term online measures could undermine long-term effectiveness.

Digital did not make strategy unnecessary, but it made execution unusually absorbing. There was always another campaign to launch, platform to learn, audience to optimise, dashboard to update or conversion point to improve. Marketing teams increasingly organised themselves around channels and systems, and professional expertise became closely associated with the ability to operate them.

Over time, channel choices began to crowd out customer choices. Optimisation displaced diagnosis. A media plan was described as a marketing strategy, and a collection of platform tactics was mistaken for a coherent approach to growth.

This was not because digital marketing failed. It was because it offered so many useful and measurable things to do.

The urgent gradually displaced the important.

AI could repeat the same mistake

The early AI conversation follows a familiar pattern. Marketers are being encouraged to identify the best tools, write better prompts, automate more workflows, produce more content and deploy agents to complete increasingly complex tasks.

These are useful capabilities. They can reduce repetitive work, shorten production cycles and make expertise more accessible. But they remain execution capabilities.

They help answer the question, “How can we do this?”

They do not necessarily answer, “Should we do this at all?”

AI can generate hundreds of content ideas without knowing whether the market needs more content. It can create message variations without determining which idea the brand should consistently own. It can summarise customer research without deciding which customers the business should prioritise or which it is willing not to serve. It can produce a campaign without taking responsibility for the commercial reasoning behind it.

The research on the “jagged technological frontier” demonstrates why this distinction matters. In the Boston Consulting Group experiment, AI produced substantial performance gains on tasks that fell within its capabilities. On a task deliberately designed to sit outside that frontier, however, consultants using AI were 19 percentage points less likely to reach the correct answer than those working without it.

The implication is not that AI should be avoided. It is that the technology’s value depends on the quality of the task, the context in which it is used and the judgement of the person directing and evaluating the work.

AI can make valuable work faster. It can also make poorly chosen work faster.

If marketing responds to AI by producing more activity without improving the decisions behind it, the profession will repeat the mistake it made during the digital era, only at far greater speed.

A product does not create its own demand

The belief that a good product will naturally find its market is persistent, particularly among founders and product-led organisations.

Occasionally, it happens. It is not a reliable growth strategy.

Customers do not evaluate every available product and select the objectively superior option. They operate with limited time, incomplete information and varying levels of interest. They rely on memory, familiarity, trust, recommendations and what is easiest to find or justify.

This is especially important in B2B markets, where purchases are often infrequent, involve several stakeholders and carry professional or organisational risk. Most potential buyers are not actively looking for a solution at any particular moment.

Professor John Dawes’ 95:5 rule is a useful way of understanding this. The exact ratio differs by category and should be treated as a heuristic rather than a universal constant. The underlying principle is that, for many B2B purchases, the large majority of potential buyers are out of market at any given time. Advertising and other forms of brand building therefore work largely by creating and refreshing memories that can be retrieved when a buyer eventually enters the category.

This has profound implications for a newly created business. Most of its future customers are not currently searching for it. They may already have a supplier, lack budget, be bound by a contract or simply have no immediate need.

Demand-capture activity can help a brand compete for the minority already in the market. Search advertising, comparison content, review platforms and sales outreach can all play an important role at that stage. But they do not explain how the brand entered the buyer’s mind before the search began.

Someone has to create the memory. The customer must encounter the brand, understand what it stands for and associate it with a relevant need or buying situation. Familiarity and trust have to be established before the organisation appears on a shortlist.

Research from the Ehrenberg-Bass Institute describes this through mental and physical availability. Brands grow by becoming easier for more category buyers to think of and easier for them to buy. That requires reach beyond the small group already shopping, as well as consistent connections between the brand and the situations that bring people into the market.

This is the work that happens long before a sale is visible in a dashboard. It is also the work that many emerging businesses underestimate because its effects are less immediate than the act of launching a product.

From the first visitor to a scalable business

AI can help a business create its first website, but it cannot guarantee the first visitor. Once that visitor arrives, the challenge becomes understanding what brought them there, whether the offer is relevant and what would persuade them to return.

An audience is not merely accumulated traffic. It is a group of people who recognise value in what a business consistently provides. Developing one requires a clear understanding of the people the organisation wants to serve, the problems it can credibly help solve and the ideas it can contribute better than the alternatives.

Turning that audience into customers requires more than visibility. The offer must be clear, appropriately priced and easy to evaluate. The customer must believe the value outweighs the cost and risk of changing. In B2B markets, the proposition may also need to survive scrutiny from procurement, finance, IT, legal and executive stakeholders.

Scaling introduces another set of questions. The business must determine which acquisition channels are repeatable, whether customer economics are sustainable, whether the brand can extend beyond the founder’s personal reputation and whether the proposition remains distinctive as competitors respond.

These are not separate from product development. They are part of building the business.

A technically scalable product with no repeatable route to market is not yet a scalable company. A large potential audience that cannot be reached economically is not yet a viable market. High engagement that does not translate into customer value or revenue is not yet growth.

Marketing connects the idea to the economic reality around it.

When execution becomes abundant, judgement becomes scarce

The falling cost of execution does not make marketing judgement less important. It makes the difference between execution and judgement easier to see.

Execution turns an instruction into an output. Judgement determines whether the instruction is sound.

Judgement is required to distinguish an interesting idea from a commercially valuable one. It is needed to interpret incomplete or contradictory evidence, to understand the trade-offs between short-term revenue and long-term growth, and to choose between several plausible courses of action.

It is also required because marketing decisions are rarely made in a controlled environment. Customers change, competitors react, budgets are constrained and organisations bring their own history, capabilities and politics to the decision. The same evidence can lead to different choices depending on the market, business model and appetite for risk.

AI can contribute to that process. It can retrieve information, challenge assumptions, identify patterns, develop scenarios and improve the speed at which a marketer explores a problem. It may even recommend a course of action.

But the marketer remains responsible for deciding whether the evidence is credible, whether the recommendation fits the context and which consequences the organisation is prepared to accept.

This is why stronger technology increases the need for stronger foundations. The more powerful the tool, the more important it becomes to know what good marketing looks like.

Evidence informs. Judgement decides.

AI should give marketers time to become marketers again

There is a more constructive opportunity in this shift.

For years, marketers have argued that they do not have enough time for strategic work. Their days are consumed by campaign administration, reporting, presentations, data cleaning, asset production, platform management, stakeholder requests and repetitive coordination.

AI can reduce some of that burden. The evidence from writing, coding and professional knowledge work suggests that well-designed AI assistance can return meaningful time to people while improving performance on suitable tasks.

The important question is what organisations and marketers choose to do with that time.

The easiest response is to demand more output. If a marketer can produce five pieces of content in the time previously required to produce one, the organisation may ask for five. If reporting can be automated, it may request more dashboards. If campaigns can be launched more quickly, it may increase the number of campaigns.

That would capture the productivity gain without necessarily improving marketing.

The greater opportunity is to reinvest the time in work that has been crowded out by execution: talking to customers, studying the category, understanding competitors, improving the offer, clarifying positioning and identifying the real barriers to growth. Marketers could spend more time working with product teams on what should be built, with finance on how value should be priced and with sales on how the market actually buys.

They could make better briefs, ask harder questions and stop activity that does not serve the strategy. They could think more carefully about how brands are built in memory, how future demand is created and how the organisation can become easier to choose.

In that sense, AI should not simply make marketers faster at the version of the job they have inherited.

It should give them the capacity to reclaim the version of marketing the business needs.

Returning to first principles

Returning to first principles is not an argument for resisting technology or retreating to an earlier era of marketing. Digital platforms, analytics, automation and AI have all expanded what the profession can accomplish. The task is to reconnect those capabilities to the purpose they are meant to serve.

First principles provide a stable foundation when the tools are changing quickly. They require marketers to begin with the market rather than the channel and with the customer rather than the campaign. They ask whether a real need exists, whether the organisation can create meaningful value, whether the offer is distinctive, and whether customers can easily think of it, find it and buy it.

They also give marketers a better way to evaluate new technology. The important question is not simply whether a tool can generate content, automate a workflow or reduce cost. It is whether it helps the organisation understand customers more clearly, make a better decision, create greater value, build stronger memories or improve the customer’s ability to buy.

Technology should strengthen marketing capability rather than substitute for marketing thought.

Without first principles, a business can execute efficiently in the wrong direction. With them, AI becomes more useful because it is being applied to better-defined problems, with stronger evidence and clearer standards for evaluating the outcome.

Five questions to ask before you build

If the ability to create is no longer the constraint, the discipline is in what you ask before you create. Before the next product, course, newsletter or campaign, answer these honestly:

1. Do enough people experience this problem to sustain a business, and is it important enough that they would change their behaviour to solve it?

2. Are they dissatisfied with the current alternatives, or merely aware that alternatives exist?

3. Can you reach them economically, through channels that are repeatable rather than lucky?

4. Why would they trust an unfamiliar provider, and what will earn that trust before the first sales conversation?

5. What will make them think of you when they finally enter the market, given that most of them are not buying today?

If you cannot answer these, the product is not ready, no matter how quickly it can now be built. If you can, most of your marketing strategy has already written itself.

The return of marketing

The AI era will create more builders. It will allow individuals and small teams to develop products that would once have required substantial organisations. It will reduce the distance between an idea and its execution and make experimentation cheaper.

This is a meaningful democratisation of capability.

It will not democratise attention in the same way.

Customers will still have limited time and more choices than they can properly evaluate. Businesses will still need to decide which markets to enter and which problems are worth solving. New products will still need to become known, remembered, trusted and easy to buy. Organisations will still need to balance the demand they can capture today with the demand they must create for tomorrow.

Those challenges cannot be solved by production alone.

They require marketers who understand the commercial role of the discipline, who can distinguish activity from progress and who know how to use evidence without surrendering judgement. They require marketing teams that contribute before the product is complete, not simply after the launch date is set.

FP Collectiv exists to help restore that version of marketing.

We believe marketing is a commercial discipline, not a promotional service. It begins with understanding markets and customers, creating value and making choices about where and how a business can grow. Communication matters, but it cannot repair a weak proposition or create lasting demand for something the market does not value.

We believe digital and AI should increase the contribution marketers make to a business. They should reduce lower-value work, improve access to evidence and give marketers more time to think. They should not trap the profession in an ever-accelerating cycle of content, campaigns and optimisation without a clear strategic foundation.

Above all, we believe that first principles become more valuable as the world becomes more complex. Tools will change. Channels will change. The economics of production will change. The need to understand people, create value, build demand and make sound commercial decisions will not.

Everyone can now build.

Few know how to market.

That is the gap FP Collectiv exists to close.

Evidence informs. Judgement decides.

Ready to build the judgement, not just the output? FP Collectiv’s structured courses teach marketing from first principles: evidence-led, commercially grounded, and sequenced so AI lands on top of fundamentals rather than instead of them. Start with B2B Marketing Fundamentals → B2B Marketing Fundamentals

New here? Subscribe for the next edition.

Subscribe now

Sources

• American Marketing Association, Definitions of Marketing, on marketing as the creation, communication, delivery and exchange of value.

• Philip Kotler, The Past, Present and Future of Marketing, on the development of marketing and the centrality of customer value.

• Shakked Noy and Whitney Zhang, Experimental Evidence on the Productivity Effects of Generative Artificial Intelligence, Science, 2023.

• Fabrizio Dell’Acqua et al., Navigating the Jagged Technological Frontier, Harvard Business School and Boston Consulting Group field experiment.

• GitHub, Quantifying GitHub Copilot’s Impact on Developer Productivity, controlled experiment examining task-completion speed.

• Les Binet and Peter Field, The Long and the Short of It, IPA analysis of short-term activation and long-term brand building.

• Ehrenberg-Bass Institute for Marketing Science, Advertising Effectiveness and the 95:5 Rule, on out-of-market buyers and the role of advertising in building memory.

• Ehrenberg-Bass Institute for Marketing Science, research on how brands grow through mental and physical availability.