Marketing has been optimising the smaller number for a decade. New IPA Effectiveness data: ROI is 11% of what makes campaigns work. Budget is 89%. The size problem we don't admit.
Marketing has a measurement obsession with a particular flavour: we love efficiency, distrust scale, and quietly pretend the two are interchangeable. They aren't. The latest data from the IPA Effectiveness Databank suggests we've been optimising the wrong number for the better part of a decade. Peter Field and Les Binet stood on stage at the IPA Effectiveness Conference 2025 with a slide that should make every CMO mildly nauseous.
Since COVID, ROI on marketing has risen 4%. Real, inflation-adjusted profit from marketing has fallen 11%.
That isn't a contradiction. That's the trick.
The maths problem nobody puts on a slide
Effectiveness, measured in profit, is the product of two things: how cleverly you spend (ROI) and how much you actually spend (budget). The industry has decided one of those things is interesting and the other is the finance department's problem.
Ask CMOs which one matters more for the outcome. They'll tell you 65% comes from spending smartly, 35% from spending big. A reasonable split for a craft-respecting profession. Slightly off-message for reality.
The IPA ran the actual regression on hundreds of campaigns. The numbers came back like this:
Budget: 89% of the variation in effectiveness.
ROI: 11%.
Or, less politely: how much you spend is roughly 8 times more important than how cleverly you spend it. Field reckoned he'd say 9 before someone with a calculator gently corrected him, which is the kind of detail you only get from a man who has earned the right to round up.
Why budget beats brain
The reason is unflattering, but obvious once you sit with it. ROI is a narrow distribution. Most well-run campaigns sit in a similar range, and most of what variation does exist isn't under your control anyway. Budget is an enormous distribution. It runs from nothing to infinity, and it's almost entirely under your control. The variable that moves most, and that you can move most, drives the outcome most.
This is the part where the room tends to go quiet.
The death spiral nobody admits they're in
What happens when marketing budgets fall? In the UK, the advertising-to-sales ratio has dropped 40% from its pre-pandemic baseline. That money didn't move into smarter channels. It just left.
The replacement logic was predictable. Cut budget. Refocus on the 5% of the market currently in-buying-mode. Call it "performance". Watch short-term efficiency rise. Watch the long-term sales base erode. Cut budget again, because the long-term sales base just eroded. Field calls this the death spiral, and he isn't being theatrical.
The numbers say this is now the default operating mode for half the brands in the IPA database. They aren't bad marketers. They're well-trained marketers, running a doctrine that was sold to them as best practice, optimising the variable that turns out to matter least.
Three moves that go the other way
If the diagnosis is "we've become too small in our ambitions", the prescription is mildly uncomfortable.
Spend bigger. The IPA's threshold for a statistically detectable lift in long-term sales is somewhere between 30 and 60 million impressions. Three Instagram posts and a TikTok partnership do not meet this threshold. They never have. Selling the CFO on "we'll need real money to move the needle" is harder than selling them on "we'll be efficient with what we have", but the second pitch produces the slow death of the brand.
Spend broader. Digital is excellent at conversion. It is mediocre at reach. The case studies that win Effies are almost always multi-channel - online plus TV plus outdoor plus radio - because audiences scatter and you have to meet them where they already are. Laithwaite's, a UK wine retailer, ran a heavy-performance machine for years and added a small brand TV layer almost as an experiment. Revenue moved by millions. That isn't magic, that's geometry: bigger audiences make bigger numbers.
Fewer ads, bigger ads, better ads. Modern brands burn 60-70% of their working budget on production - 14 different cuts per campaign, bespoke assets for every channel, the agency timesheet eating the media plan. Replace it with one big idea, produced once, deployed everywhere, repeated until your internal team is sick of it. Then keep going, because the audience hasn't even noticed yet. Wear-out is mostly a myth invented by people who watch their own ads too many times.
The contrarian bit
Here's the observation that's been quietly sitting under all this. The thing most CMOs are best at - making the existing budget go further - is the thing that matters least. The thing most CMOs are worst at - arguing for a bigger budget in front of a CFO - is the thing that matters most. The skill set most rewarded is the one most decorative. The skill set least rewarded is the one most valuable.
That's why effectiveness work feels so strange when you encounter it for the first time. It tells you the boring answer is the right answer. Spend more, on fewer ideas, in bigger places, for longer. Don't get cute. Don't optimise the dashboard. Build a thing that thirty million people remember in two years.
Go big, in other words, or go home.
