Almost every store that gets going hits the same wall somewhere between £20k and £40k a month. Spend goes up, revenue moves a little, and margin quietly goes backwards. The wall is rarely where owners think it is, and it is almost never solved by spending harder.
Almost every store that gets going hits the same wall somewhere between £20k and £40k a month. Spend goes up, revenue moves a little, and margin quietly goes backwards. The wall is rarely where owners think it is, and it is almost never solved by spending harder.
The Growth Marketing Agency
·
28 September 2026
A store gets to about £20k a month on one channel that works, usually paid social or Shopping, sometimes a single strong product. Then growth flattens. The instinct is to spend more, because spending more is what got you here. That is exactly when margin starts leaving.
A store gets to about £20k a month on one channel that works, usually paid social or Shopping, sometimes a single strong product. Then growth flattens. The instinct is to spend more, because spending more is what got you here. That is exactly when margin starts leaving.
The ceiling is real, it is predictable, and it has a small number of causes. This piece names them, gives you the three numbers that tell you which one you have, and sets out the order to fix them in.
The ceiling is real, it is predictable, and it has a small number of causes. This piece names them, gives you the three numbers that tell you which one you have, and sets out the order to fix them in.
The three numbers that tell you where you actually are
The three numbers that tell you where you actually are
Most stores are flying on revenue and platform ROAS, which is why they cannot see the wall coming. These three tell you the truth instead.
Most stores are flying on revenue and platform ROAS, which is why they cannot see the wall coming. These three tell you the truth instead.
Contribution
Contribution
After every fee
Revenue minus cost of goods, minus payment and platform fees, minus shipping and returns, minus ad spend. The only number that tells you whether growth is worth having.
Blended CAC
Blended CAC
Not channel CAC
Total marketing spend divided by total new customers. Platform-reported CAC always looks better than this one, and this one is what your bank balance responds to.
Repeat rate
Repeat rate
At 90 and 365 days
What proportion of customers buy again. A store with a 15% repeat rate and a store with a 45% repeat rate can afford completely different acquisition costs.
Where £100 of store revenue actually goes
Where £100 of store revenue actually goes
A representative mid-margin store scaling on paid. Yours will differ, which is exactly the point of working it out.
£38
£28
£9
£4
£21
Cost of goods
The part everyone models correctly.
Ad spend
At a 3.5x blended return. Push for growth and this is the segment that grows.
Shipping and returns
Free delivery thresholds and a 12% return rate.
Payment and platform fees
Small, fixed, and almost never in the target ROAS.
Contribution
What is actually left to pay for the business.
At this shape, a drop from 3.5x to 2.8x blended return does not trim contribution, it roughly halves it. That is why scaling on revenue rather than contribution goes wrong so quickly.
The quick diagnostic
Take your last three months. If revenue grew and contribution did not, the ceiling is in your economics. If contribution grew but total orders did not, the ceiling is in your demand. If neither moved while spend went up, the ceiling is in your site. Those three cases need completely different work, and doing the wrong one is how a year disappears.
The shape of a stall
The shape of a stall
Monthly ad spend against monthly revenue, indexed to month one.
M1
M2
M3
M4
M5
M6
M7
M8
M9
Ad spend
Ad spend
Revenue
Revenue
Revenue is still going up, which is why this is hard to spot from a revenue chart. But from month five each extra pound of spend is buying less, and contribution has been falling since month four.
The seven reasons stores stall
The seven reasons stores stall
REASON 01
One channel carrying everything
One channel carrying everything
Most stores at £20-30k are one channel deep. That channel has a natural ceiling of addressable audience, and once you pass it you are paying more for worse buyers. It is also a single point of failure the day the account gets restricted.
REASON 02
The offer never changed
The offer never changed
The product that got you to £20k was competitive against a smaller set of competitors. Bundles, thresholds, guarantees and pricing architecture are levers most stores touch once and never revisit, and they move contribution faster than any bid strategy.
REASON 03
Discovery is broken above a few hundred SKUs
Discovery is broken above a few hundred SKUs
Navigation, filtering and on-site search decide whether a visitor finds the thing they came for. On a growing catalogue this quietly becomes the biggest leak on the site, and no amount of traffic fixes it.
REASON 04
Checkout is losing people nobody counts
Checkout is losing people nobody counts
Shipping cost revealed late, no guest checkout, a payment method your buyers expect and you do not offer. This is the most expensive part of the journey and usually the least examined.
REASON 05
Margin has not been modelled properly
Margin has not been modelled properly
Returns, payment fees, pick and pack, shipping subsidies and discounting rarely make it into the target ROAS. Stores scale confidently on a number that was never right, and the losses arrive at the end of the quarter.
REASON 06
Retention treated as a nice-to-have
Retention treated as a nice-to-have
Email and CRM are the cheapest revenue in the business and usually the last thing resourced. A store that lifts repeat rate can afford to outbid its competitors on acquisition indefinitely.
REASON 07
Tracking that stopped being true
Tracking that stopped being true
Consent changes, iOS, ad blockers and a Shopify theme update. Most stores are optimising to a signal that degraded eighteen months ago and nobody re-checked.
Why spending harder makes it worse
Why spending harder makes it worse
Within a channel, the cheapest, most intent-heavy audience gets bought first. As you scale, each additional pound reaches someone less likely to buy. Your average cost per acquisition rises even though nothing about your ads got worse.
Within a channel, the cheapest, most intent-heavy audience gets bought first. As you scale, each additional pound reaches someone less likely to buy. Your average cost per acquisition rises even though nothing about your ads got worse.
“Doubling spend on a channel that is already at its efficient ceiling does not double revenue. It buys you worse customers at a higher price and hides the problem for another quarter.”
“Doubling spend on a channel that is already at its efficient ceiling does not double revenue. It buys you worse customers at a higher price and hides the problem for another quarter.”
THE SCALING TRAP
Return on the last £5,000, not the average
Return on the last £5,000, not the average
Blended return produced by each additional £5,000 of monthly spend on a single channel.
4.1x
First £5k
3.4x
to £10k
2.7x
to £15k
1.9x
to £20k
1.3x
to £25k
0.9x
to £30k
Average return across the whole account might still read 2.4x here and look perfectly healthy. The marginal return is what tells you the last two increments were bought at a loss.
The number that matters here is marginal, not average. If the last £5,000 you added produced a materially worse return than the £5,000 before it, you have found your ceiling for that channel. That is a signal to widen the mix or fix the economics, not to push harder.
The number that matters here is marginal, not average. If the last £5,000 you added produced a materially worse return than the £5,000 before it, you have found your ceiling for that channel. That is a signal to widen the mix or fix the economics, not to push harder.
The order to fix it in
The order to fix it in
This order matters. Doing them out of sequence is how stores spend a year and end up in the same place.
This order matters. Doing them out of sequence is how stores spend a year and end up in the same place.
Fix tracking first, because every decision after this depends on the numbers being real.
Fix tracking first, because every decision after this depends on the numbers being real.
Model true contribution per product and per order, including returns and fees. Some of your best sellers are probably losing money.
Model true contribution per product and per order, including returns and fees. Some of your best sellers are probably losing money.
Fix the site where it leaks: discovery, product pages, then checkout. This is free traffic you have already paid for.
Fix the site where it leaks: discovery, product pages, then checkout. This is free traffic you have already paid for.
Work the offer: bundles, thresholds, guarantees. This lifts contribution without touching media spend.
Work the offer: bundles, thresholds, guarantees. This lifts contribution without touching media spend.
Turn on retention properly. Flows before campaigns, and measure revenue per recipient rather than open rate.
Turn on retention properly. Flows before campaigns, and measure revenue per recipient rather than open rate.
Only then widen the channel mix, funded by the margin the first five steps released.
Only then widen the channel mix, funded by the margin the first five steps released.
Why this order releases cash
Steps one to five cost media budget almost nothing and improve the return on every pound you were already spending. They also raise the ceiling on what you can afford to pay for a customer, which is what makes step six affordable. Stores that start at step six and skip the rest are the ones that stall again at £60k.
What good looks like on the other side
What good looks like on the other side
A STORE AT THE CEILING
One channel producing 80% or more of revenue
Reports on revenue and platform ROAS
Target ROAS set by rule of thumb
Email is a monthly newsletter
Returns and fees looked at quarterly
A STORE PAST IT
Three or four channels, each with a defined job
Reports on contribution after every fee
Target set from modelled margin per product
Flows running, measured on revenue per recipient
Contribution per order visible weekly
None of this is exotic. It is unglamorous work on economics, site and retention, done in the right order, before the budget goes up rather than after. That is the whole difference between a store that stalls at £30k and one that does not.
None of this is exotic. It is unglamorous work on economics, site and retention, done in the right order, before the budget goes up rather than after. That is the whole difference between a store that stalls at £30k and one that does not.
Want to know which ceiling you are hitting?
Want to know which ceiling you are hitting?
Want to know which ceiling you are hitting?
The free strategy works through your own numbers: contribution after fees, where the ceiling actually sits, and what to do about it in what order. Yours to keep either way.
The free strategy works through your own numbers: contribution after fees, where the ceiling actually sits, and what to do about it in what order. Yours to keep either way.