Marketing ROI: The Formula Most Teams Are Using Wrong

Marketing ROI measures the return generated by marketing activity relative to its cost. The basic formula is straightforward: subtract marketing costs from the revenue attributed to marketing, divide by marketing costs, and multiply by 100 to get a percentage. What makes it hard is not the arithmetic. It is deciding what counts as revenue, what counts as cost, and whether the attribution holding the whole thing together is honest.

Most teams calculate it correctly on paper and get it wrong in practice. The formula is not the problem. The assumptions feeding it are.

What Is the Basic Marketing ROI Formula?

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

So if you spent £50,000 on a campaign and it generated £200,000 in attributed revenue, your marketing ROI is 300%. You made £3 back for every £1 you spent, net of the original investment.

Some teams use a simpler ratio, often called ROAS (return on ad spend), which skips the subtraction and just divides revenue by spend. A 4:1 ROAS means £4 of revenue for every £1 spent. ROAS is useful for campaign-level optimisation. ROI is the right metric when you are making budget decisions at a business level, because it accounts for the cost of the activity, not just the output.

Both are valid. Neither is reliable unless the attribution underneath them is honest, which is where most of the real work sits.

What Should You Include in Marketing Costs?

This is where teams regularly undercount, which inflates ROI in ways that flatter the marketing function but mislead the business.

Marketing costs should include media spend, agency fees, technology and platform costs (your CRM, your marketing automation, your analytics stack), content production, internal headcount directly attributed to marketing activity, and any overheads that sit within the marketing function. If you run a campaign and your in-house design team spends three weeks on it, that time has a cost. Leaving it out makes the campaign look more efficient than it was.

I have seen this done selectively throughout my career, usually not through dishonesty but through habit. Teams report on what they track, and headcount rarely sits inside the marketing budget line. The result is ROI numbers that look strong but would collapse if you ran them on full cost. When I was running agencies, I saw clients present campaign ROI to their boards using media spend alone as the cost base, with agency fees, technology subscriptions, and internal resource all sitting in separate budget lines. The maths looked great. The actual return on total investment was considerably less impressive.

A clean ROI calculation captures all the inputs, not just the ones that make the output look good.

How Do You Handle Attribution When Calculating Marketing ROI?

Attribution is the single biggest variable in any marketing ROI calculation, and it is almost always the least scrutinised.

Most digital marketing platforms default to last-click attribution. That means the final touchpoint before a conversion gets 100% of the credit. If someone saw your brand campaign three times, clicked a retargeting ad, searched for your brand name, and then converted through a paid search ad, that paid search ad gets the full revenue attribution. The brand campaign, the retargeting, and the organic search experience get nothing.

This systematically overstates the ROI of lower-funnel channels and understates the ROI of upper-funnel activity. It is one of the reasons performance marketing budgets tend to grow and brand budgets tend to shrink, even in businesses where brand investment is doing most of the commercial work.

Earlier in my career I was guilty of leaning too hard on lower-funnel performance metrics. The numbers were clean, the attribution was direct, and the ROI looked excellent. What I came to understand over time is that much of what performance marketing gets credited for was already going to happen. You are often capturing intent that already existed rather than creating demand that would not have existed without you. The retargeting ad that “converted” someone who had already decided to buy is not generating ROI in any meaningful sense. It is collecting credit for someone else’s work.

Better attribution models, including linear attribution, time-decay models, and data-driven attribution where you have enough volume, spread credit more honestly across the customer experience. None of them are perfect. All of them are more useful than last-click for understanding what your marketing is actually doing.

For a broader look at how attribution and measurement fit into growth strategy, the Go-To-Market and Growth Strategy hub covers the commercial frameworks that sit behind these decisions.

Should You Calculate ROI by Channel or Across All Marketing Spend?

Both. But you should trust them differently.

Channel-level ROI is useful for optimisation decisions within a channel. It tells you which campaigns, audiences, or creative executions are performing better relative to spend. That is operationally valuable. The problem is that channel-level ROI is almost always distorted by attribution, because channels report on their own performance using their own measurement logic, and every channel tends to claim more credit than it deserves.

Blended marketing ROI, which divides total revenue growth attributable to marketing by total marketing investment, is a more honest commercial signal. It is harder to game because you are not relying on any single platform’s attribution model. If you spent £500,000 across all marketing activity in a year and can credibly attribute £2.5 million in incremental revenue to that activity, your blended ROI is 400%. That number is less precise than channel-level figures, but it is more honest.

The challenge with blended ROI is the word “incremental.” You need to separate the revenue that marketing generated from the revenue that would have come in anyway through existing customers, word of mouth, and baseline demand. That requires some form of incrementality testing or econometric modelling, neither of which is trivial. But the alternative, treating all revenue as marketing-attributed, produces numbers that are flattering and meaningless.

Tools like those covered in Semrush’s analysis of growth tools can support measurement and optimisation at the channel level, but they do not resolve the attribution question. That requires judgement, not software.

What Is a Good Marketing ROI?

It depends on the business model, the margin structure, the competitive environment, and what you are measuring. There is no universal benchmark that means anything.

A 5:1 revenue-to-spend ratio is often cited as a reasonable target for marketing ROI, meaning £5 of revenue for every £1 spent. In some industries with high margins and strong brand pricing power, that is conservative. In low-margin retail or highly competitive categories, it may be unachievable without cutting corners on attribution honesty.

What matters more than hitting a specific number is whether your ROI is improving over time, whether it holds up under scrutiny, and whether it is being calculated consistently so you can compare periods meaningfully. A business that calculates a 200% ROI using full costs and honest attribution is in a better position than one reporting 600% ROI using partial costs and last-click attribution, even though the second number looks better in a board presentation.

I spent several years judging the Effie Awards, which measure marketing effectiveness rather than creative merit. One thing that struck me repeatedly was how rarely entrants could clearly articulate the commercial return on their campaigns, even when the work was genuinely impressive. The industry has become very good at measuring activity and very inconsistent at measuring outcomes. That gap is where ROI calculations tend to fall apart.

How Do You Calculate Marketing ROI for Brand Activity?

This is the genuinely hard problem, and anyone who tells you they have solved it cleanly is oversimplifying.

Brand marketing builds awareness, preference, and trust over time. Its effects are diffuse, delayed, and difficult to isolate. You cannot run a brand campaign in January and point to a specific revenue line in February and say that is the return. The causal chain is real but it is long, and it runs through customer perception, not through a trackable click.

The most credible approaches to measuring brand ROI include marketing mix modelling (econometrics), which uses statistical analysis to decompose revenue growth into its contributing factors over time. It is expensive to do well, requires significant historical data, and is not accessible to smaller businesses. For those without the budget or data for econometrics, brand tracking surveys, share of search as a proxy for brand health, and controlled market tests (running brand activity in some regions but not others and comparing outcomes) can provide directional evidence even if they cannot produce a precise ROI figure.

The honest answer for most businesses is that brand ROI can be approximated but not precisely calculated. That does not make brand investment unjustifiable. It means you need to make the case for it on a different basis: the long-term commercial logic of building a brand that customers prefer, that commands better pricing, and that requires less performance spend to convert. Forrester’s work on intelligent growth models captures some of this logic, particularly the relationship between brand health and sustainable revenue growth.

The mistake is either ignoring brand ROI because it is hard to measure, or pretending you can measure it with the same precision as a paid search campaign. Both positions are wrong.

How Do Time Horizons Affect Marketing ROI Calculations?

Most ROI calculations are too short. They measure what happened in the campaign window or the fiscal quarter, which captures the immediate return but misses the compounding effects of marketing investment over time.

A customer acquired through a brand campaign in Q1 might not convert until Q3. A piece of content published today might drive qualified traffic for three years. A loyalty programme that reduces churn creates value that accumulates over the lifetime of each retained customer, none of which shows up in a 30-day ROI window.

Customer lifetime value (CLV) is the metric that corrects for this. If your average customer spends £200 in their first year but £800 over their lifetime, calculating ROI on first-year revenue alone understates the return on acquisition by a factor of four. Businesses that optimise for short-term ROI without accounting for CLV tend to underspend on acquisition and overspend on lower-funnel activity that captures quick conversions but does not build the customer base that drives long-term growth.

When I was growing the agency from a small team to over 100 people, one of the commercial decisions we had to make was how to value new client relationships. A client worth £50,000 in year one might be worth £300,000 over a five-year relationship if we retained and grew them well. Calculating ROI on acquisition costs against year-one revenue would have made most of our new business investment look marginal. Calculating it against projected lifetime value changed the decision entirely. The same logic applies to marketing ROI in any business with meaningful customer retention.

This connects to a broader point about how growth strategy should be structured. The Go-To-Market and Growth Strategy hub goes into the frameworks for thinking about acquisition, retention, and long-term commercial value in more depth.

What Are the Most Common Mistakes in Marketing ROI Calculations?

Beyond the attribution and cost-base issues already covered, there are a few consistent errors worth naming directly.

The first is confusing correlation with causation. Revenue went up while you were running a campaign, so the campaign gets the credit. But revenue might have gone up anyway due to seasonality, a competitor’s problems, a PR moment, or a product change. Isolating what marketing actually caused requires some form of control, whether that is a holdout test, a geographic split, or an econometric model. Without it, you are measuring coincidence.

The second is optimising for ROI efficiency rather than ROI scale. A campaign that returns 800% ROI on a £10,000 budget is less commercially interesting than a campaign that returns 300% ROI on a £500,000 budget, because the absolute return is larger. Teams that chase efficiency metrics can end up running very small, very optimised programmes that do not move the business. This is particularly common in performance marketing, where the best-performing audiences are often the smallest ones closest to conversion. Semrush’s examples of growth approaches illustrate how scale and efficiency interact differently depending on the channel and business model.

The third is using ROI as a justification rather than a diagnostic. When ROI is calculated after the fact to validate a decision already made, it tends to produce the number needed to make the case. When it is used as a genuine diagnostic to understand what is working and why, it produces something more useful: a clearer picture of where to allocate the next pound of marketing spend.

The fourth, and perhaps the most underappreciated, is treating marketing ROI as a standalone metric disconnected from the underlying business. Marketing ROI is partly a function of product quality, pricing, customer service, and brand reputation, none of which marketing controls entirely. I have worked with businesses where the product genuinely delighted customers and marketing ROI was strong almost regardless of what tactics were used. I have worked with others where no amount of marketing efficiency could compensate for fundamental product or service problems. If your ROI is persistently low, the question worth asking is whether the problem is the marketing or the business it is supporting. BCG’s research on go-to-market strategy makes this point in the context of financial services, but the principle applies broadly: commercial performance reflects the whole business, not just the marketing function.

How Do You Build a Marketing ROI Framework That Actually Works?

Start with the business question, not the metric. What decision are you trying to make? If you are deciding how to allocate next year’s budget across channels, you need blended ROI with consistent cost accounting. If you are optimising a live campaign, you need channel-level efficiency metrics. If you are making the case for a brand investment, you need a longer-term model that connects brand health to commercial outcomes over time.

Then define your revenue attribution approach before you run the activity, not after. Agree on what counts as marketing-attributed revenue, which attribution model you will use, and what the measurement window will be. Changing these definitions after the fact to improve the numbers is the most common way ROI calculations lose credibility internally.

Build in a baseline. What revenue would you have generated without the marketing activity? This does not need to be a complex econometric model for every campaign. Even a rough estimate, based on prior periods with similar conditions, is better than attributing all revenue to marketing by default.

Report on full costs, including headcount, technology, and agency fees, not just media spend. This will make your ROI numbers look less impressive in the short term and more credible over time.

And treat the number as a starting point for a conversation, not an endpoint. A 250% ROI is useful information. What is more useful is understanding which parts of the activity drove it, which parts did not, what you would do differently, and whether the result is repeatable at scale. BCG’s work on go-to-market planning emphasises the importance of building measurement into the launch strategy from the start rather than retrofitting it. That principle holds whether you are launching a pharmaceutical product or a digital campaign.

Marketing ROI is not a number you find. It is a number you build, with assumptions that need to be honest, costs that need to be complete, and attribution that needs to reflect reality rather than flatter the function doing the measuring.

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