The 12-point ad account checklist
These are the twelve things we check on every Meta and Google Ads account before we touch a budget, in the order we check them. Nothing gated, nothing withheld — work through it yourself if you want to.
The order matters more than the list. Measurement comes first because every later decision is made with the numbers it produces, and optimising against wrong data is the most expensive mistake in paid media. Economics come second, because you cannot judge a campaign without knowing what a customer is worth. Only then is it worth arguing about structure.
Measurement — do this first
One primary conversion action per platform
What good looks like: Exactly one action marked primary, and it corresponds to revenue or a qualified lead.
If it is wrong: With several primary actions the bidding optimises toward the cheapest one, which is almost never the valuable one.
Server-side events with deduplication
What good looks like: Conversions API on Meta and enhanced conversions on Google, with a shared event ID so browser and server events are recognised as one.
If it is wrong: Without deduplication your conversions are inflated and the algorithm is learning from noise.
Event match quality above 6
What good looks like: Meta Events Manager shows 6.0 or better on Purchase, ideally 8 and up.
If it is wrong: Low match quality means Meta cannot confirm who converted, so it bids cautiously and underspends.
Platform numbers reconciled against the store
What good looks like: You know the gap between what the platforms claim and what your store recorded, and why it exists.
If it is wrong: If Google and Meta added together exceed your actual revenue, both are claiming the same customers.
Economics — before any budget decision
Contribution margin per order is known
What good looks like: Product cost, shipping, fees, discounts and returns all subtracted. A real number, not a guess.
If it is wrong: Without it, every ROAS target is arbitrary and you cannot tell a profitable campaign from a busy one.
Break-even ROAS written at the top of reporting
What good looks like: 1 divided by your contribution margin as a fraction. A 30% margin means 3.3x.
If it is wrong: Campaigns below break-even lose money on every sale, however good the dashboard looks.
Blended MER tracked monthly
What good looks like: Total revenue divided by total ad spend, taken from the store, not the platforms.
If it is wrong: If platform ROAS rises while MER flatlines, you are buying customers you already had.
Structure and delivery
Enough conversions per ad set to exit learning
What good looks like: Roughly fifty conversions per ad set per week. No ad set stuck on 'learning limited'.
If it is wrong: Budget spread across too many ad sets starves all of them and the system never bids confidently.
No audience overlap above about 25%
What good looks like: Lookalikes and interest stacks checked against each other in the overlap tool.
If it is wrong: Your own ad sets bid against each other and push your costs up.
Brand terms separated from Performance Max
What good looks like: A brand exclusion list on PMax and a dedicated brand Search campaign with a tight bid cap.
If it is wrong: PMax reports a flattering ROAS by intercepting people who were already searching for you.
Search terms reviewed in the last 30 days
What good looks like: Ninety days exported, sorted by spend, with account-level negative lists for jobs, DIY and research queries.
If it is wrong: In most accounts 20 to 45% of search spend is on queries that were never going to buy.
Where the traffic lands
Landing page LCP under 2.5s on a mid-range phone
What good looks like: Field data in PageSpeed Insights, on mobile, not the lab score on your laptop.
If it is wrong: You pay full price for the click and lose the visitor before the page renders.
We run all 12 on your account for $99
Same checklist, run by a senior media buyer on your actual accounts, with the wasted spend totalled and a fix list ordered by impact. Three business days, and the fee comes off your first invoice if you go on to work with us.
See the ad account auditFind out what your account is really doing
A free fifteen-minute scan: we look at one ad account and tell you the three biggest problems we can see, in writing — whether or not you hire us.