A premium travel brand
100+ campaigns consolidated to 10, ROAS from 0.67x in April 2025 to 6.11x in November 2025, on 58% less spend.
Founder, Marvellous Performance
Ten years of growing businesses on the numbers their owners actually care about: bookings, deposits, subscriptions and revenue, not clicks.
years growing businesses
countries: Nigeria, Ghana, Kenya, Uganda, the UK and the US
Up to
a month in Google Ads budgets
At most agencies, the senior people win the account and juniors run it. Here, our founder oversees every account, and a senior account manager runs it day to day.
That means the person who set your strategy is the one watching whether it works. When something needs changing, it doesn’t wait for a junior to notice and escalate.
Our founder
Oversees every account.
The person who set your strategy is the one watching whether it works.
A senior account manager
Runs it day to day.
When something needs changing, it doesn't wait for a junior to notice and escalate.
The problem
Operating in 28 metros, it had to win more first-order customers with the budget frozen. Platform-reported numbers were no longer accepted: only deduplicated customers the ads actually caused counted, and payback had to stay under six months.
What changed
One definition of a new customer across Google and Meta, counted once. Budget moved weekly by marginal return, and spend was timed to the months when demand was high and ads were cheap.
Incremental new customers
Cost per incremental customer
Payback: about 4.5 months, every month inside the six-month limit.
The result, Year 1 to Year 2, on the same $14.69 million
Lower is better.
| Measure | Year 1 | Year 2 | Change |
|---|---|---|---|
| Incremental new customers | 418,879 | 503,767 | +20.3% |
| Cost per incremental customer | $35.07 | $29.16 | −16.9% |
Payback about 4.5 months, every month inside the six-month limit.
The problem
The dashboards looked strong. Restated on what the ads actually caused, after returns and fulfilment, profit was far lower, and Meta was losing money once existing customers and view-through were stripped out.
What changed
Spend on warm audiences and existing customers cut back where it lost money at the margin, new-customer prospecting protected, and Google scaled where the next dollar still returned more than it cost. Halo effects, such as brand search, were measured separately and never counted towards the goal.
Incremental contribution after ads
New-customer share of orders, from 50.2%
Budget, the same both years
The result, Year 1 to Year 2, on the same $50.4 million
| Measure | Year 1 | Year 2 | Change |
|---|---|---|---|
| Incremental contribution after ads | $10.72m | $12.39m | +15.5% |
| New-customer share of orders | 50.2% | 53.1% | up |
Meta was still losing money at the end, but less: from −$2.57 million to −$1.75 million, while still bringing in most of the new customers.
Nigeria, Ghana, Kenya, Uganda, the United Kingdom and the United States. E-commerce, travel, fintech, SaaS, mobility and education.
Two books on the method: First 90 Days of Growth, and Google Ads Mastery.
If you'd like your account run by the people who've done this before, start with a conversation about your numbers.