Better ads at a flat budget: cost per order €84.38 → €58.63
Commentary byROASArchiveOur analysis, not theirsAn ecommerce brand kept its budget flat, focused on stronger ads and cut weak ones, and cost per order fell from €84.38 to €58.63 (−30.5%). @SamPiliero's team runs the account.
Orders rose from 279 to 393 over the last 30 days, while spend dipped slightly.
- Cost per order: €84.38 → €58.63 (−30.5%). Each order cost much less.
- Orders: 279 → 393 (+40.9%), with slightly lower spend.
- Total sales: €29,182.71 → €36,389.04 (+24.7%). Revenue grew less than orders.
- AOV: €104.60 → €92.59 (−11.5%). A weak point: baskets got smaller, which the team puts down to product mix.
- New customers: 274 → 385 (+40.5%). Almost all orders came from first-time buyers.
The change
Before
In the earlier 30 days, spend was €23,541, total sales were €29,182.71 and the account got 279 orders at €84.38 each.
After
The team spent more effort on fewer, stronger ads, cut low performing mid-funnel ads, and built new rounds from the winners. Spend was €23,040.
Our reading: why it likely worked
Our interpretation, not the advertiser's claim.
- More ads only help when they are good enough to earn spend. Focusing on quality gave the algorithm better options.
- Cutting weak mid-funnel ads stopped budget leaking into ads that didn't convert, while the best top-of-funnel ads kept bringing new buyers.
- Building each round from past winners stacks small gains over time.
Caveats: This is a before/after over two 30-day periods on one account, not a split test. Three changes were made at once, so we can't say which one mattered most, and the falling AOV means revenue gains were smaller than order gains.
How to run this in your account
- Hypothesis
- If we produce fewer but stronger ads, cut low performers and iterate winners, then cost per order will fall because spend shifts to ads that convert.
- Primary metric
- Cost per order
- Guardrails
- AOV
- New customer share
- Total sales
- Who should try it
- Accounts with many ads and flat results, where weak ads take a share of spend. It is less useful for small accounts that only run a few ads.
Questions
Can you lower CPA without raising budget?
This account did. Spend fell slightly from €23,541 to €23,040, and cost per order dropped from €84.38 to €58.63.
Should I cut mid-funnel ads on Meta?
This team cut weak mid-funnel ads but kept the best converters and top-of-funnel ads.
They say it is tricky, so cut based on results, not funnel stage alone.
Why did AOV drop when results improved?
AOV fell 11.5% to €92.59, which the team links to product mix. More orders still lifted total sales 24.7%.