📖 Glossary

Marketing ROI: What It Is, How to Measure It Honestly & The Evidence

Marketing ROI is the incremental profit from marketing as a percentage of its cost. Here's the honest way to calculate it, what the IPA evidence shows, and the traps to avoid.

·6 min read

Definition (≤25 words). Marketing ROI is the incremental profit generated by marketing activity, expressed as a percentage of the marketing investment that produced it.

TL;DR. Marketing ROI is the financial return on marketing spend — the number that answers "did this pay for itself?" Done honestly, it measures incremental profit over the right time horizon, and it distinguishes short-term activation returns from the larger, slower, long-term brand returns. Done badly — last-click, quarterly, profit-blind — it systematically misleads and quietly starves the brand building that drives most of the value. ROI is the currency of marketing effectiveness; the question is whether you're calculating it with discipline or with wishful thinking.

Why does ROI matter more than every metric on your dashboard?

Every marketing dashboard in the world is cluttered with metrics that measure activity — impressions, reach, clicks, engagement rate, cost-per-lead, ROAS. These are useful for operational management. None of them, individually, tells you whether the marketing is making the business richer. ROI is the metric that answers the P&L question: for every pound spent, how much incremental profit came back?

This matters because marketing budgets are negotiated with finance directors, not with marketing directors. When the conversation turns to "what did we get for the £4 million we gave you?", a defensible ROI answer keeps the budget alive; a shrug gets it cut. More deeply, ROI discipline forces a choice: spend on activity that generates commercial return, or spend on activity that generates activity. Without ROI as the organising principle, marketing drifts toward whatever is measurable on a dashboard — which, as we'll see, is not the same as what is effective.

The distinction from efficiency matters here and matters everywhere. Return on ad spend (ROAS) measures revenue attributed to ad spend. It says nothing about profit (you can have high ROAS on a loss-making product), nothing about incrementality (you can have high ROAS on demand that would have arrived anyway), and nothing about the long term. ROAS is an efficiency metric. ROI is an effectiveness metric. Confusing the two is the single most expensive error in performance marketing.

How do you calculate it honestly?

The formula is simple. The honesty is not.

ROI (%) = (incremental profit attributable to marketing − marketing cost) ÷ marketing cost × 100

Two words do all the heavy lifting: incremental and attributable.

Incremental means the profit that would not have existed without the marketing. A brand with strong organic demand will sell product regardless. If your marketing targets people who were going to buy anyway, you haven't generated profit — you've harvested it at a cost. Incrementality is established through causal methods: control groups, holdout geographies, and geo-experiments. These are the gold standard. Marketing mix modelling (MMM) — econometric modelling of historical sales data — is the aggregate alternative when holdouts aren't feasible.

Attributable means crediting the marketing that actually caused the sale, not the touchpoint that happened to be last. Last-click attribution — the default in most digital platforms — credits the final interaction before purchase. It systematically over-credits search and retargeting (which sit at the bottom of the funnel catching already-decided buyers) and systematically under-credits the brand advertising that put the brand in the buyer's mind in the first place. This isn't a minor distortion. It is a structural bias that, left uncorrected, pushes budgets toward the channels that take credit and away from the channels that create demand.

What does the evidence show about short-term and long-term ROI?

The most important finding from Les Binet and Peter Field's analysis of the IPA Effectiveness Databank — an archive of over a thousand campaigns with proven commercial return — is that advertising produces two kinds of return on two different clocks.

Short-term ROI comes mostly from activation: price promotions, direct response, targeted digital. It shows up fast — within days or weeks — and is relatively easy to measure. But it is small in absolute terms and decays quickly. A price promotion lifts sales while it runs; when it stops, sales revert. Worse, habitual promotion erodes price elasticity — customers learn to wait for the deal.

Long-term ROI comes mostly from brand building: emotional, broad-reach advertising that grows penetration, mental availability, and pricing power. It pays back over months and years, not weeks. And here is the crucial point: Binet and Field's data consistently show that the majority of total profit return from effective campaigns accrues over the long term — the short-term spike is real but represents a minority of the total value.

This is why the 60:40 rule exists: because if you only measure short-term ROI, you will conclude that activation is your best investment and brand building is waste. You will cut brand spend, your short-term ROI will look better for a quarter or two, and then your baseline will begin to erode as mental availability decays and pricing power weakens. This is the pattern Binet and Field describe as the downward spiral of short-termism: each quarter's dashboard looks better; the business gets steadily weaker.

Where people get it wrong

  • Confusing ROI with ROAS. ROAS measures attributed revenue per pound of ad spend. It ignores profit, incrementality, and the long term. A campaign with strong ROAS on a low-margin product promoted to customers who would have bought anyway may have negative ROI. Optimising ROAS in isolation is how brands quietly destroy value.
  • Optimising for high ROI rather than absolute return. ROI is a ratio. A small bottom-funnel retargeting campaign generating a modest absolute return can show a handsome ROI percentage. A large brand campaign can show a lower ratio but vastly more pounds in the bank. Chasing the ratio starves the scale. Always pair ROI with absolute return.
  • Measuring everything on a quarterly clock. Activation ROI is legitimately a short-term metric. Brand ROI is not. Demanding that brand building prove itself within 90 days is like demanding that a sapling produce fruit in a season — you'll cut down every tree before it has a chance to grow. Match the measurement window to the job.
  • Trusting last-click attribution as ROI. It is not ROI; it is an attribution model with a known structural bias toward bottom-funnel channels. Use it as a diagnostic, never as the sole basis for budget allocation.
  • Treating "not attributable" as "didn't work." The most valuable marketing — brand building that grows share of search and penetration — is the hardest to attribute. Absence of attribution is not absence of effect.

In practice — what ROI discipline actually looks like

  1. Set the time horizon by objective. Activation: weeks to a quarter. Brand: 6 to 24 months. Never judge brand ROI through a quarterly lens.
  2. Measure incrementality with control groups wherever possible. Geo-holdout tests are the practical gold standard. MMM is the alternative when holdouts are not feasible.
  3. Report both short-term and long-term ROI separately. A campaign that looks mediocre at three months may be the best investment you've made at eighteen.
  4. Correct for last-click bias. Use MMM or media-mix experiments to validate — or overturn — what the attribution platform tells you. The gap between attributed and incremental is often where the real story lives.
  5. Pair ROI with absolute return and scale. Report the pounds generated, not just the ratio. A board needs to know how much money came back, not just how efficient the spend was.
  6. Make ROI the organising metric — not the only metric. Use it to adjudicate between investments at the portfolio level, but keep the diagnostic metrics (reach, frequency, share of search) for operational management.

So what?

Marketing ROI is the number that keeps marketing funded — and the number that, calculated badly, quietly defunds the best work marketing does. The discipline is not in the formula; it's in the two hard words: incremental and attributable. Measure only the short term and you will optimise toward a channel mix that looks efficient and erodes the business. Measure honestly — incrementality, the right horizon, both clocks — and you'll find that the unglamorous, slow, hard-to-attribute brand work is almost always where the real money is. That's not a convenient truth for quarterly dashboards. It is the evidence.

Evidence & sources

  • Binet, L. & Field, P. (2007). Marketing in the Era of Accountability. IPA.
  • Binet, L. & Field, P. (2013). The Long and the Short of It. IPA.
  • Binet, L. & Field, P. (2018). Effectiveness in Context. IPA.
  • IPA Effectiveness Databank — the underlying archive of proven-ROI cases.
  • Sharp, B. (2010). How Brands Grow. Oxford University Press — penetration and mental availability as the mechanisms behind long-term brand value.
  • WARC — ROI benchmarking and effectiveness case data.

Definition type: DefinedTerm. Reviewed: 2026-07-09. Written by Andrei Tarnovski for Boringly Effective. Fundamentals over fashion.