The death spiral

Les Binet's analysis of the IPA Effectiveness Databank, presented with Will Davis at the IPA Effectiveness Conference in October, should have redrawn every marketing budget in the country. Budget, they found, explains 89% of the variation in profit generated by award-winning campaigns. ROI explains just 11%. Yet when 500 senior marketers were surveyed, 65% said ROI mattered more. Six months later, the marketing industry's principal response to this finding has been to build more tools for measuring ROI.

To be clear about what Binet means by these two things: ROI is the efficiency of your spend, the return you get per pound invested. Budget is the total amount you invest. His research says that how much you spend matters almost nine times more than how efficiently you spend it. A £5m campaign with mediocre efficiency will almost always outperform a £500,000 campaign with outstanding efficiency, because scale creates reach, reach creates fame, and fame is what drives long-term business growth.

The 2025 martech landscape catalogued 15,384 marketing technology solutions, up from 150 in 2011. Measurement tools are among the fastest-growing categories within that. Open-source marketing mix models, no-code scenario planners, AI-powered budget allocators — what once required a dedicated econometrician and a six-figure contract is now available to teams with modest budgets and no data science function. The language around all of this is democratisation. On paper, it should be good news, but it has not played out that way.

Forrester's 2026 B2C marketing predictions forecast a 7% decline in marketers' confidence in their ability to measure business impact. A Harvard Business Review Analytic Services report published on 27 March found that nearly 40% of marketers struggle to connect marketing mix model outputs to real-world business decisions. According to research published by the Trade Desk in February, only 22% say they can confidently justify their value to their CFO with the data they currently have. Nikki Taylor, a former CMO at UPS now working in the measurement industry, put it more bluntly earlier this year: 'I was fluent in the language of marketing ROI, but I had a secret. For most of that time, I didn't actually know what was really driving business impact.'

The tools have multiplied, the data has deepened, and confidence has fallen. The CMO Survey puts a number on the gap between what marketers know they should do and what they actually do: the ideal budget split, according to marketing leaders themselves, is 50% long-term brand building and 50% short-term performance. The actual split is 31% and 69%.

The reason that gap persists has nothing to do with the quality of the tools. It is baked into how organisations use them.

Measurement exists, in most companies, to answer a question posed by finance: prove this works before we give you more money. It is a defensive instrument. The marketing department produces dashboards and attribution reports and ROAS figures not because these things guide strategy, but because they justify the budget that already exists. The better the dashboard looks, the safer next quarter's spend.

Binet's research says something uncomfortable about this arrangement. If budget is the dominant variable in marketing effectiveness, then the single most important thing a marketer can do is secure a larger budget. But the measurement apparatus is designed to prove efficiency within the current one. It optimises the 11% while ignoring the 89%. You cannot prove the impact of the investment you were never given, and no marketing mix model in the world can backtest a scenario in which your CFO said yes to a budget they actually said no to.

This is the trap. The tools work. They do exactly what they were designed to do, which is to make incremental spend accountable. They were never designed to make the case for transformational investment, and the more sophisticated they become, the more they anchor the conversation around efficiency rather than scale.

This did not arrive overnight. The past 15 years of digital marketing have been a sustained push toward channels and tactics that can be precisely measured: programmatic buying, search, social, retail media, the entire performance marketing ecosystem. All of it was built on the promise that marketing could finally be made accountable. And it was. But as budgets migrated toward the measurable, the work that was hardest to attribute — brand campaigns, sponsorship, broad-reach advertising, the things that build fame over years — was gradually defunded. Not because it stopped working, but because it could not compete in a quarterly review against a paid search campaign with a clean ROAS number. The bill for that trade-off is now arriving.

Binet and Davis put hard numbers on what that trade-off has cost. Since 2018, media ROI across IPA cases has increased by 4%. Incremental profit has fallen by 11%. Marketers are getting better returns on less money and calling it progress, while their actual contribution to business growth shrinks. The dashboards have never looked better, but market share tells a different story.

The cycle reinforces itself: tighter targeting, smaller budgets, higher reported ROI, lower actual profit. A marketer who delivers a 5:1 ROAS on a £500,000 spend generates £2.5m in revenue. A marketer who delivers 2:1 on £5m generates £10m. In a quarterly review, the first marketer looks like the smarter operator. In the P&L, the second one is four times more valuable to the business.

It is worth asking who benefits from this arrangement. The measurement industry is not a neutral observer. Every vendor in the space — from the platforms that offer free tools to the consultancies that sell proprietary ones — has a commercial interest in making the channels they can measure look effective. Google's Meridian and Meta's Robyn are the most visible examples: free, open-source marketing mix models built by companies whose primary business is selling advertising inventory. But the incentive runs through the entire ecosystem. Attribution vendors, ad tech platforms, retail media networks — each one is calibrated to make its own contribution legible and measurable. Channels that are harder to attribute, such as outdoor, sponsorship, or long-term brand campaigns, tend to show up less clearly in the output of almost any model, regardless of who built it. The marketers most exposed to this are mid-budget teams without a statistician to interrogate model quality, who take the output at face value because the interface is clean and the recommendations feel precise.

Nike and Airbnb are the high-profile examples that get cited here, and neither maps neatly onto a mid-market brand. But the principles are worth noting. Under its previous CEO, Nike increased its advertising budget while shifting spend toward performance channels and measurable digital activity. There was no shortage of data. Over four years, the company lost $28bn of market value. His replacement, Elliott Hill, has since told investors he is shifting dollars from performance marketing back to brand. At Airbnb, the trajectory ran the other way: the company cut performance marketing spend from 2019 onwards, redirected investment toward brand, and by 2022 had reported its first full-year profit on revenues of $8.4bn, with total marketing spend down 28%. In both cases, the defining decision was made on conviction, not on the output of a model. The measurement followed.

The tools are not the problem. The problem is what they are used for. In most organisations, measurement is a defensive instrument: proof that last quarter's spend was justified.

The most useful thing a marketer can do with their measurement stack right now is turn it into an offensive one. Take the same scenario planning tools that are currently used to optimise channel mix within a fixed budget and use them to model what happens with a bigger budget. Ask a different question: not 'where should we allocate this spend?' but 'what would a 30% increase in total investment deliver?' Build the case for scale using the tools that were designed to argue for efficiency. That is not a misuse of measurement. It is the highest-value use of it.

The evidence is already there. Budget is eight times more important than ROI. The brands that grow are the ones that spend enough. The only thing missing is the willingness to put that evidence in front of finance and argue for a different size of bet.

Ross Farquhar, marketing director at Little Moons, captured the tension well: 'I've been in organisations that don't really care about the data and just take wild leaps based on opinion. Sometimes that's brilliant, sometimes it's shaky ground. At the other end, I've seen businesses where the investment in effectiveness creates an industry in itself. You spend a disproportionate amount of time and money measuring rather than putting things in front of the consumer. Neither is healthy.'

He is right. And the answer is not to abandon measurement. It is to stop using it as a shield and start using it as a lever. Not 'can we prove this worked?' but 'what happens if we spend enough to find out?'
‍

Looking for a new agency partner? DCA cuts through the noise to connect brands with agencies that fit

• We cover all types of projects and disciplines, from strategy and creative to performance and tech.
• Our editorial team vets every agency, so you skip the pitch theatre and get straight to capable partners.
• We offer consultancy, pitch management, or a complimentary shortlisting service if that's all you need. Submit a brief today

‍