Four offer-selection filters — unit economics, network reputation, fit with proven combinations, and a disciplined test — that filter out roughly 70% of offers before a euro is spent.
Behind every successful Ads Sign campaign is a selection process that filters out roughly 70% of the offers under consideration before the first test. It is not a formality — it directly determines whether the media buying team spends its time scaling or putting out fires.
Unit economics on paper, before launch
We calculate expected CPA from historical vertical and GEO data and compare it against the payout: if the model doesn't add up even in the optimistic scenario, the offer never reaches testing, no matter what the network manager promises.
Advertiser or network reputation
Approval rate, payout speed and stability, history of hold disputes. Information is pulled from an internal database of past launches and from closed buyer chats — the numbers in the network dashboard rarely tell the whole story.
Fit with proven combinations
A new offer in a familiar vertical with a well-understood audience clears the process faster than an offer that looks more profitable on paper but sits in a vertical where the team has no proven expertise.
Test launch with a hard decision deadline
Offers that clear the first three filters go into a test launch with a capped budget and a hard decision deadline — typically 48-72 hours to reach a statistically significant volume of data, not "let's watch it for another week."
Scaling decisions by data, not by gut feel
If the test confirms the economics, budget is increased in stages with continuous ROI monitoring; if it doesn't, the offer is shut down without regret, even if time has already gone into preparing creatives for it.
In short
- The offer's unit economics must add up on paper before launch.
- Network reputation is checked against an internal database and buyer chats, not just the network dashboard.
- A familiar vertical with proven expertise clears selection faster.
- The test is capped by budget and time — 48-72 hours to a decision.
- The test decides scaling or shutdown, not the buyer's gut feel.
There's no universal recipe here, but the mistakes listed above cost the most and repeat the most often.