The wrong question, measured brilliantly

Modern B2B marketing can model a buyer's intent from behavioural signals. It can run multi-touch attribution across dozens of touchpoints. It can score which accounts are in-market and debate the relative merits of algorithmic versus rule-based attribution weighting. What it usually cannot do is answer one piece of arithmetic: did we recover more from this customer than it cost to acquire them?

That question is not epistemologically complex. It needs gross margin, acquisition cost, and a cohort of customers. And yet, in most B2B businesses, it has never been answered with any rigour, while a great deal of effort has gone into questions that are far harder. The sophistication is real. It has simply been pointed in the wrong direction. This is the story of how that happened, and it is worth saying at the outset that nobody designed it on purpose.

01A model borrowed from a different business

Digital channels created trackable data where none had existed before, and for a while that felt like unambiguous progress. In some contexts it genuinely was. A consumer e-commerce brand running paid search can trace a click to a basket to a purchase in minutes. There is no sales force in the middle, the decision is made by one person, and the sale resolves inside the measurement window. The metrics are commercially real because the thing they measure actually happens, at scale, in the time the measurement covers.

B2B businesses saw this and adopted the model. The adoption was understandable, and structurally mistaken. B2B has a sales force between the marketing activity and the closed deal. The cycle runs weeks or months. The buyer is usually a committee, and the decision involves procurement, legal and executive sign-off. Apply consumer measurement logic to that commercial motion and the metrics that look most impressive are the top-of-funnel ones: clicks, leads, cost per lead, conversion from ad to form. They respond quickly to campaign changes and they produce tidy numbers for a report. What they do not capture is the value of what happens next, through the sales process, over the months that follow. The measurement window closes before the commercial story resolves.

The metrics that look best in B2B are the ones that resolve fastest. They are rarely the ones that decide whether the spend was worth it.

02Tools built for someone else's motion

The marketing technology industry grew up in the same period, and it was built, naturally enough, as software-as-a-service, because that was the dominant delivery model of the era. More to the point, many of these tools were built by businesses whose own commercial motion was inbound and product-led. HubSpot is the canonical example: a company that developed and proved the inbound model for a self-serve, trial-led motion, and whose tools reflected those assumptions.

B2B businesses with complex enterprise sales cycles adopted those tools anyway, because the tools existed, looked sophisticated, and promised the accountability that B2B marketing had always struggled to demonstrate. The instrument was designed for one kind of buying behaviour and deployed against another. That mismatch did not announce itself. It simply meant the numbers that came out most easily were, again, the fast ones at the top.

03Instrumented in the image of its operators

A marketing operations function grew to run these platforms, and an agency industry grew alongside it offering the same expertise at scale. MarOps practitioners are typically skilled in data integrity, process governance, attribution modelling and technical platform management. These are genuine and important capabilities, and the point that follows is not a criticism of the people who hold them.

What was less present in the formation of the function was commercial marketing knowledge: an understanding of how acquisition cost connects to customer lifetime value, how the ratio between the two determines whether marketing is creating or destroying capital, and how to configure a measurement infrastructure to answer that question. The tools were instrumented in the image of their operators. What came out reflected what went in.

The agencies do good work, and answer real questions about pipeline visibility, attribution and process efficiency. The difficulty is not the quality of their answers. It is that the questions they answer are all second questions. The first question, whether marketing is recovering more from each customer than it cost to win them, is not in the engagement scope. It never has been.

04When the framework failed, the answer was more framework

To its credit, the industry noticed that something was not working. Adoption of the technology fell sharply: Gartner found that utilisation dropped from 58% of available capabilities in 2020 to 33% by 2024.¹ McKinsey, surveying the same landscape, described rampant duplication across technology stacks and little demonstrable return at the organisational level after years of investment.²

The response was not to simplify, or to ask a more fundamental question. It was to add more instrumentation. Intent-signal platforms emerged, designed to capture behavioural indicators that a buyer might be active: content downloads, website visits, search patterns, third-party data signals. Account-based marketing, presented as a new discipline, is in practice a formalisation of what good B2B salespeople have always done: build relationships with specific accounts. IBM was doing this at scale fifty years ago. It did not have a name then. It was simply how complex business was conducted.

Intent modelling deserves a direct word, because it is the clearest example of the pattern. It is the industry's attempt to close the gap between the slow, relationship-driven B2B motion and the fast, trackable consumer model it had borrowed. In B2C, at scale, with millions of data points and near-instant feedback, behavioural proxies for purchase intent have reasonable validity. In B2B, with small account populations, long and largely invisible buying journeys, and decisions made in conversations that leave no digital trace, the proxy is much weaker. You can instrument every touchpoint on your website and still have no visibility of the conversation the procurement director had with your competitor at an industry event last week. That conversation is the signal that matters, and it is not capturable by any platform.

05The observation that ties it together

Here is where the sophistication landed, and it is worth stating plainly rather than dressing up.

Marketing cannot reliably say whether it recovers more from one customer than it costs to acquire them. That is simple arithmetic. And the same function, simultaneously, deploys intent-signal platforms, builds probabilistic models of buyer behaviour, runs multi-touch attribution across dozens of digital touchpoints, and holds considered debates about attribution weighting.

The sophistication went towards the questions that are hardest to interrogate, and away from the one that decides whether marketing creates value. A complex framework is easier to hide inside than a simple ratio.

It is not hard to see why, once you stop looking for villains. A discipline that has long struggled to demonstrate commercial accountability will tend to prefer the question it can answer impressively over the one it might answer badly. Complexity is a form of cover. None of it requires bad faith, and most of the people involved were rational actors responding to the incentives and the tools in front of them.

06Why this is a B2B problem in particular

The people who are best placed to challenge all this, the CFOs, the investment partners, the CEOs who control the budget, have been handed a language they cannot easily interrogate and metrics they cannot independently verify. The result is visible in how senior leaders now regard the function. Forrester reports that 64% of B2B marketing leaders do not trust their own measurement to make decisions.³ The Boathouse CEO study found that 60% of chief executives now view marketing as a cost centre, up from 35% the year before.⁴ When the people holding the budget stop believing the numbers, the numbers were never doing the job they were built to do.

This is, to a degree, an inside problem, and an honest account should say so. The path into B2B marketing has often run through adjacent skills rather than formal commercial training, and the discipline has tended to default to activity, content and campaigns, to metrics that measure the activity rather than its effect. That is an observation about how the profession developed, not a charge against individuals. But it does mean the measurement gap is structural, persistent, and expensive, rather than a temporary oversight that better tooling will fix.

07The simpler question, asked first

The way out is not another platform. It is a return to the question the whole apparatus has been built around without ever answering: across a cohort of customers, does the fully loaded cost of winning them come back, with margin, over their life? Calculated to a standard a finance team would sign off, by segment rather than blended, that single ratio tells you what no amount of intent modelling can: whether the marketing is creating value or quietly consuming it.

It is a less impressive question than the ones the industry has learned to ask. It is also the only one that settles the argument. The dashboards can stay. They are simply answering the second question. This is the first.