You gave us a wonderfully simple answer to a wonderfully simple question: for every £1 we spent, how much came back? But journeys got complicated, growth got expensive, and somewhere along the way the metric started making the decisions.
You've been there through countless years of Google Ads reports, monthly performance meetings and awkward conversations about why Meta had a better month than forecast. You gave us a wonderfully simple way of answering a wonderfully simple question: “For every £1 I spent, how much revenue did I get back?”
And for a long time, that was enough. But things have changed. We want more.
We want to meet new customers. We want to grow. We want to invest in channels that introduce people to the brand before they're ready to buy. And, increasingly, we're realising that you might be holding us back. ROAS, we think we've outgrown you.
First, an important clarification. ROAS is not bad. It remains an incredibly useful metric for understanding the efficiency of advertising spend. If you spend £100 and generate £500 in attributed revenue, knowing you've achieved a 5x ROAS is useful information.
The problem starts when ROAS stops being ‘a metric’ and becomes ‘THE business model’.
Because ROAS looks at paid media largely in isolation. It asks what revenue can be attributed back to the money you spent on advertising. What it doesn't adequately answer is what that advertising did to everything else:
This is your friend and mine, ‘the halo effect of marketing’, and it's one of the areas where a strict ROAS model starts to struggle. A customer journey rarely looks like:
SEE AD → CLICK AD → BUY PRODUCT
It might look more like:
SEE INSTAGRAM AD → IGNORE IT → SEE ANOTHER AD → GOOGLE THE BRAND → READ SOME REVIEWS → COME BACK DIRECTLY → BUY
So which channel gets the credit? More importantly, should that determine whether the original advertising was worthwhile? Good old attribution conversations. We've all been there.
If your sole objective is hitting a strict ROAS target, advertising platforms have a fairly obvious incentive. Find the people most likely to buy. And who are often the people most likely to buy? People who already know your brand.
Existing customers. Previous website visitors. People searching directly for you. High-intent audiences who are already close to making a purchase. They're fantastic customers to advertise to. They're also finite.
Eventually, if a business wants to grow, it has to convince people who don't already know it exists to become customers. And those customers are, nine times out of 10, more expensive to acquire. And that's where the relationship with ROAS gets complicated.
Imagine your target is 5x ROAS. Your established audience happily delivers 5x, 6x or 7x. But reaching completely new customers delivers 3x initially. If the rule is simply “we need 5x ROAS”, the answer is obvious: “Turn it off”.
Congratulations. You've protected your ROAS. Unfortunately, you may also have just turned off your next generation of customers.
This is perhaps the biggest reason brands should reconsider how heavily they rely on ROAS. A business can become “more efficient while becoming less capable of growth”.
If you continually optimise towards the easiest customers to convert, your ROAS might look fantastic. But your pool of customers isn't necessarily getting any bigger. There is a natural tension between “maximising today's efficiency” and “creating tomorrow's demand”.
Growth requires experimentation. It requires entering new audiences. It requires reaching people earlier in their buying journey. It sometimes requires accepting that the 10,001st customer will cost more to acquire than the 1,001st. A rigid ROAS target gives marketers very little room to do that.
Unit economics sounds considerably less exciting than ROAS. It also sounds like something you're going to need a spreadsheet for, and you'd be right. But the principle is actually very simple. Instead of asking “How much revenue did my advertising generate?” we start asking:
“What is a customer worth to my business, and how much can I afford to spend acquiring one?”
That changes the conversation. Suddenly we're thinking about things like:
These are business metrics rather than advertising metrics. And that's important. Because Google doesn't pay your salaries. Meta doesn't pay your warehouse costs. Your ROAS doesn't go into your bank account.
Profit does.
Imagine two campaigns.
| Campaign A | Campaign B | |
|---|---|---|
| Revenue | £100,000 | £200,000 |
| Media spend | £20,000 | £50,000 |
| Platform ROAS | 5x | 4x |
| Verdict on a strict 5x target | “Lovely stuff” | “Turn it off” |
Illustrative figures, for the maths rather than any specific account.
If your target is 5x ROAS, Campaign B looks like the problem. But what if the business has sufficient margin to acquire those additional customers profitably at 4x? Campaign B has generated an additional £100,000 of revenue while reaching considerably more customers. If those customers buy again, recommend the brand, search for it later or purchase through organic channels in the future, the value becomes greater still.
So the question shouldn't necessarily be “Did we hit 5x ROAS?” It should be: “Was that additional £30,000 of investment profitable, and did it help us grow?” That's a much more interesting question.
Moving towards unit economics establishes the real economic boundaries marketers can operate within. If we know how much gross profit an average order generates, how frequently customers return and what a customer is worth over time, we can calculate how much we're genuinely willing to spend acquiring them.
That gives marketing teams room to distinguish between different types of investment. Perhaps we're willing to pay considerably more for a brand-new customer than someone who purchased last month. Perhaps a first purchase only breaks even because we know a meaningful percentage of those customers purchase three more times. Perhaps introducing 50,000 new people to the brand increases paid acquisition costs today but creates more branded search, direct traffic and organic revenue six months from now.
ROAS alone struggles to accommodate those conversations. Unit economics encourages them.
Absolutely not. We're not blocking ROAS's number and pretending the relationship never happened. ROAS remains a useful diagnostic metric, and we're big believers in a sitewide view of ROAS. The change is recognising that ROAS should inform decision-making rather than dictate it.
Use ROAS to understand advertising efficiency. Use CAC to understand what it costs to acquire customers. Use contribution margin to understand whether those customers are profitable. Use lifetime value to understand what they're worth beyond their first transaction.
And look at blended business performance to understand whether your marketing ecosystem is actually growing. That really matters. If paid media spend increases 30%, platform ROAS falls from 5x to 4x, but total business revenue increases 40%, new customer acquisition accelerates and contribution profit grows significantly… was marketing performance really worse?
Your ROAS dashboard might say yes. Your bank account might disagree, and that's where the owners, and our friends in the finance team, are always most interested.
You don't need to ceremonially delete ROAS from every dashboard on Monday morning. No need to send it packing. If working with REBEL, we'd start by understanding a few fundamental numbers:
WHAT DOES AN AVERAGE CUSTOMER SPEND? → HOW MUCH GROSS PROFIT DOES THAT GENERATE? → HOW OFTEN DO THEY PURCHASE AGAIN? → WHAT ARE WE PREPARED TO PAY TO ACQUIRE A NEW CUSTOMER?
From there, marketing targets can start being built around the economics of the business rather than an arbitrary efficiency ratio. You can still monitor ROAS. You just stop asking it to make every decision.
ROAS helped digital marketing grow up. It gave marketers accountability and connected advertising investment directly with commercial outcomes. But marketing has become more complicated. Customer journeys span channels. Attribution is imperfect. Paid advertising influences organic behaviour. Platforms increasingly make their own optimisation decisions. And businesses need to balance efficiency today with creating demand for tomorrow.
So we don't need to completely break up with ROAS. Maybe we just need to change the relationship. Keep ROAS around, see how it's doing. But when it comes to deciding how much you should invest in growth? Maybe it's time to start seeing other metrics…
This is what giving a damn looks like in practice. Tell us what's not working.