Marketing ROI Formula: Stop Measuring the Wrong Number

The marketing ROI formula is straightforward: subtract your marketing costs from the revenue generated, divide by those costs, and multiply by 100 to get a percentage. What makes it complicated is not the maths. It is deciding what counts as marketing cost, what counts as attributable revenue, and over what time period you are measuring.

Get those three decisions wrong and the formula gives you a precise answer to the wrong question. That is the version most marketing teams are running.

What Is the Marketing ROI Formula?

The standard formula is: (Revenue Attributable to Marketing – Marketing Costs) / Marketing Costs x 100.

If you spent £50,000 on a campaign and it generated £200,000 in revenue, your ROI is 300%. For every pound you spent, you got four back, three of which are return above cost.

That is the version taught in textbooks. In practice, two of the three variables in that formula are genuinely difficult to pin down, and the third is not as simple as it looks either.

Revenue attributable to marketing: how do you know which revenue marketing caused? Attribution is a contested, imperfect science. Even with sophisticated modelling, you are making assumptions about which touchpoints drove which conversions, and those assumptions carry real weight in the final number.

Marketing costs: most teams undercount these. They include media spend but forget agency fees, technology costs, internal headcount, creative production, and the time senior people spend briefing and reviewing work. The fully-loaded cost of a campaign is often 30 to 50 percent higher than the media line alone.

Time window: ROI over 30 days looks different from ROI over 12 months. For categories with long purchase cycles or strong repeat purchase behaviour, a short measurement window systematically undervalues marketing investment.

Why Most Marketing ROI Calculations Are Quietly Wrong

I have sat in a lot of agency and client-side meetings where ROI figures were presented with considerable confidence, and I have spent a fair amount of time quietly unpicking the assumptions behind them. The errors tend to cluster in predictable places.

The first is cost undercounting. A paid search team reports a 600% ROI on their channel. That looks extraordinary until you add in the agency management fee, the platform technology costs, the time spent by the in-house marketing manager overseeing the account, and the creative and landing page development costs. The real ROI might be closer to 200%. Still good. But a very different conversation.

The second is last-click inflation. If your ROI calculation is built on last-click attribution, you are giving all the credit to the final touchpoint and none to the channels that built awareness, drove consideration, or re-engaged lapsed customers. Paid search, in particular, benefits enormously from last-click attribution because it sits at the bottom of the funnel and captures demand that other channels created. That does not make paid search bad. It makes last-click attribution a misleading lens for measuring its contribution.

The third is time window gaming. Measuring ROI over the period in which a campaign ran, rather than over the period in which its effects materialise, is a common way to make campaigns look worse than they are, or better, depending on which direction serves the narrative. Brand campaigns in particular need longer measurement windows to show their full value. Cutting the window short and comparing brand ROI to performance channel ROI on the same timeframe is not an apples-to-apples comparison.

Forrester has written usefully about the questions you need to ask to improve marketing measurement, and the underlying point is consistent: the measurement framework shapes the answer you get. If you design your ROI calculation to confirm what you already believe, it will.

If you want a broader grounding in how analytics thinking connects to these questions, the Marketing Analytics hub covers attribution, dashboards, and measurement frameworks in more depth.

How to Build the Inputs Properly

Getting the formula right starts with getting the inputs right. Here is how I approach each one.

Fully-Loaded Costs

Build a cost inventory that covers every resource the campaign consumed. That means media spend, agency and freelance fees, technology and platform costs, internal time (estimated at a loaded hourly rate), creative and production costs, and any incremental operational costs the campaign generated. If a promotional campaign required additional customer service headcount to handle enquiries, that is a marketing cost.

Most marketing teams do not track internal time against campaigns. That is understandable operationally, but it means the cost side of the ROI calculation is systematically understated. For large campaigns, internal time can represent a meaningful proportion of total cost, particularly in organisations where senior people are heavily involved in approvals and revisions.

Attributable Revenue

This is harder. There is no perfect method, and anyone who tells you otherwise is selling something. The pragmatic approach is to choose an attribution model that reflects how your customers actually make decisions, apply it consistently, and be transparent about its limitations.

For businesses with short, simple purchase journeys, data-driven attribution in GA4 is a reasonable starting point. For businesses with long, complex journeys involving multiple channels and significant offline activity, you will need a more sophisticated approach: media mix modelling, incrementality testing, or a combination. GA4 has meaningful improvements in how it handles attribution compared to Universal Analytics, but it is still a model, not a ground truth.

One approach I have used in practice is triangulation: run two or three attribution models simultaneously, look at where they agree and where they diverge, and use the divergence as a signal of where your measurement is most uncertain. Where models agree, you can be reasonably confident. Where they diverge significantly, treat the figures as directional rather than definitive.

Time Window

Set your measurement window based on your category’s purchase cycle, not on reporting convenience. For e-commerce with a short cycle, 30 to 60 days is usually sufficient. For considered purchases, financial products, or B2B categories, you may need 6 to 12 months to capture the full revenue effect of a campaign.

If your business has strong repeat purchase behaviour, think about whether you want to measure first-order ROI or lifetime value ROI. A campaign that acquires customers at a loss on the first transaction can still be strongly ROI-positive if those customers have high retention rates. Measuring only the first transaction systematically undervalues customer acquisition investment.

Channel-Level vs. Portfolio-Level ROI

This is a distinction that matters more than most marketing teams give it credit for.

Blended portfolio ROI is the number you report to the board. It tells you whether marketing as a whole is generating a return above cost. That is useful for justifying the overall budget and for comparing marketing investment to other business investments.

Channel-level ROI is the number you use to make allocation decisions. It tells you which channels are working, which are not, and where additional investment would generate the best marginal return. A blended ROI of 250% can hide a paid social channel running at negative ROI while paid search runs at 800%. The blended number looks fine. The allocation is broken.

Early in my career at lastminute.com, I ran a paid search campaign for a music festival that generated six figures of revenue within roughly 24 hours of going live. The channel-level ROI was extraordinary. But that success was partly a function of the campaign running at a moment when demand already existed and the brand already had recognition. The paid search channel captured that demand efficiently. It did not create it. If we had measured only paid search ROI and concluded that paid search alone was responsible for the business outcome, we would have made poor allocation decisions as a result.

Channel-level ROI is a useful operational metric. It becomes dangerous when it is used to make strategic decisions without accounting for the interdependencies between channels.

Mailchimp’s overview of core marketing metrics is a reasonable starting point for thinking about how channel metrics connect to broader performance measurement, even if the treatment is necessarily general.

The Difference Between Marketing ROI and Business ROI

Marketing ROI measures the return on marketing investment. Business ROI measures the return on total business investment. These are related but different, and conflating them is a common source of confusion in boardroom conversations.

A marketing campaign with a 400% ROI sounds impressive. But if the product being sold has thin margins, the operational costs of fulfilling the demand are high, and the marketing-generated customers have poor retention, the business ROI of that campaign might be much lower, possibly negative. Marketing ROI that does not account for margin and lifetime value is measuring revenue efficiency, not business value.

The version of the formula that CFOs find most credible incorporates gross margin rather than revenue. Instead of measuring revenue against marketing cost, you measure gross profit against marketing cost. That gives you a figure that reflects the actual economic value marketing created, not just the top-line revenue it influenced.

The formula becomes: (Gross Profit Attributable to Marketing – Marketing Costs) / Marketing Costs x 100.

This version is harder to calculate because it requires you to know your gross margin by product, channel, and customer segment. But it is the version that survives scrutiny in a finance review. I have seen marketing teams present strong revenue-based ROI figures to boards, only to have the CFO point out that the margin profile of the products being sold meant the business was effectively paying to acquire customers at a loss. That is not a conversation you want to be surprised by.

Benchmarks and What They Are Actually Worth

Marketing ROI benchmarks circulate freely. You will see figures suggesting that a 5:1 ratio (500% ROI) is strong, that anything below 2:1 is questionable, and that exceptional campaigns can reach 10:1 or higher. These numbers are not useless, but they need significant contextualisation before they mean anything.

ROI benchmarks vary enormously by industry, channel, business model, measurement methodology, and time window. A 200% ROI for a brand campaign measured over 12 months in a considered-purchase category might represent excellent performance. A 200% ROI for a direct response e-commerce campaign measured over 30 days might be mediocre.

The more useful benchmark is your own historical performance. What ROI have your campaigns delivered in the past? What were the conditions that produced your best and worst results? What does incremental investment typically return at the margin? Those questions, answered with your own data, are more actionable than any industry average.

I spent several years judging the Effie Awards, which are explicitly focused on marketing effectiveness rather than creativity. One of the consistent observations across entries was that the campaigns with the strongest demonstrated ROI were almost always built on clear, pre-defined measurement frameworks. The teams that could show strong ROI had decided in advance what they were measuring and why. The teams that struggled to demonstrate effectiveness were often measuring whatever data was available after the fact and reverse-engineering a story from it.

Forrester’s point about black-box marketing analytics is relevant here: when you cannot explain how a number was produced, it loses credibility regardless of how impressive it looks.

Applying the Formula in Practice

Here is a worked example that reflects how I would approach this in a real business context.

A mid-size retailer runs a Q4 campaign across paid search, paid social, and email. The campaign runs for eight weeks. The headline numbers look like this:

Media spend: £120,000. Agency fees: £18,000. Technology and platform costs: £6,000. Creative production: £12,000. Internal time (estimated): £8,000. Total fully-loaded cost: £164,000.

Revenue attributed to the campaign using data-driven attribution: £680,000. Average gross margin: 42%. Gross profit attributed: £285,600.

Revenue-based ROI: (£680,000 – £164,000) / £164,000 x 100 = 315%.

Margin-based ROI: (£285,600 – £164,000) / £164,000 x 100 = 74%.

Two very different numbers. The revenue-based figure looks strong. The margin-based figure is positive but modest, and that is before accounting for any customer service, logistics, or returns costs associated with the incremental volume. The margin-based ROI is the one worth taking to the CFO, because it is the one that will survive the follow-up questions.

This kind of calculation also reveals why it matters to track costs properly. If this team had measured ROI against media spend alone (£120,000), the revenue-based ROI would have been 467% and the margin-based ROI would have been 138%. Neither figure is dishonest in isolation, but both overstate the economic return on the full investment the business made.

Building a dashboard that surfaces these figures clearly, and makes the methodology visible, is as important as the calculation itself. A well-structured marketing dashboard should make the assumptions behind the numbers accessible, not just the numbers themselves.

What to Do When ROI Is Hard to Measure

Some marketing activity resists clean ROI measurement. Brand campaigns, PR, sponsorship, content marketing, and social media presence all generate value that is difficult to trace directly to revenue. That does not mean they have no ROI. It means the measurement methodology needs to be different.

For brand activity, proxy metrics matter. Brand awareness, consideration, and preference scores can be tracked through research. Share of voice can be measured against competitors. These are not revenue figures, but they are leading indicators of future revenue, and they can be used to build a case for brand investment even when direct attribution is impossible.

Incrementality testing is one of the more rigorous approaches available. By holding out a portion of your audience from a campaign and comparing their behaviour to the exposed group, you can measure the incremental lift the campaign generated. This is not straightforward to execute, but it produces more defensible ROI estimates than attribution modelling alone, particularly for upper-funnel activity.

The honest position is that some marketing investment will always be difficult to measure precisely, and that is acceptable as long as you are transparent about it. Presenting a precise ROI figure for a brand campaign when the methodology is shaky does more damage to marketing’s credibility than acknowledging that the measurement is imperfect. Finance teams respect honesty about uncertainty more than they respect confident numbers that do not hold up to scrutiny.

MarketingProfs put it well in their piece on using web analytics effectively: the value of measurement lies in the decisions it enables, not in the precision of the numbers themselves.

The broader question of how to connect ROI measurement to budget allocation decisions, and how to use analytics to make those conversations with finance more productive, is something I cover in detail across the Marketing Analytics hub. If you are building out a measurement framework from scratch, that is a useful place to start.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions