Ask most marketing teams how their ad spend is performing and you'll get a blended ROAS number, delivered with confidence. Ask which specific audience, creative, and placement combination is actually earning that number, and the confidence usually drops. Blended metrics are comforting and almost useless — they average a handful of campaigns quietly losing money against the two or three that are carrying the whole account.
The accounts we take over rarely have a spending problem. They have an allocation problem: budget keeps flowing to what was working eight months ago, because nobody built the habit of checking whether it still is.
Start by finding what's actually carrying the account
Pull performance at the campaign level, not the account level, and sort by cost per acquisition rather than total conversions. Every account we've audited has the same shape once you do this: a small number of campaigns producing most of the efficient volume, a long tail barely breaking even, and usually one or two quietly burning budget on inflated impressions with almost nothing to show for it.
That long tail is where most wasted spend hides, and it's also the easiest to fix — because the winning campaigns already exist. The fastest ROI gain most accounts can make isn't a new campaign; it's turning off three underperforming ones and moving that budget to what's already proven.
Creative fatigue is a bigger leak than most people think
Even a strong campaign decays. The same three ad variations shown to the same audience for months will keep running, keep spending, and keep converting worse each week, because frequency climbs and the audience simply stops noticing the ad. This shows up in the data as a slow decline in click-through rate that gets blamed on "the algorithm" when it's really just an audience that's seen the same creative sixty times.
"The campaign isn't broken. The creative is just tired."
Miaoon media buying notes
The fix is a refresh cadence, not a guess: rotate creative on a schedule tied to frequency, not to whenever someone remembers to. We typically flag a variation for refresh once frequency crosses 3–4 impressions per user in a rolling window, well before performance visibly drops.
An apparel client's blended ROAS looked healthy at 2.1x, but campaign-level review showed one prospecting campaign at 4.3x carrying two retargeting campaigns running near breakeven. We shifted 35% of the retargeting budget into the top prospecting campaign and refreshed stale creative on the rest. Blended ROAS moved to 3.4x within six weeks, without increasing total spend.
A simple audit framework you can run this week
- Rank campaigns by cost per acquisition, not total conversions, to separate efficient volume from expensive volume.
- Check frequency on every active campaign. Anything climbing past 3–4 needs new creative before it needs more budget.
- Set a floor. Decide the CPA above which a campaign gets paused, and hold to it — sunk cost is the single biggest reason weak campaigns survive review after review.
- Move budget toward the winners weekly, not quarterly. Ad platforms shift fast; a review cadence that's too slow keeps money in campaigns that stopped working weeks ago.
None of this requires a bigger budget. It requires treating the budget you already have as something to actively manage, rather than something to set once and revisit at the end of the quarter.