I see this all the time in B2B SaaS accounts. Someone's running retargeting campaigns, patting themselves on the back for the CTR, and they don't realise the platform is basically spending money on cold traffic. We audited one account recently that had dropped over $60K on acquisition. Looked fine on the surface - traffic, conversions, pacing. But dig into the tracking and the whole thing was a house of cards.
Conversion events were duplicating, so the algorithm was optimising against inflated garbage. Key funnel stages were either missing or misconfigured, so nobody had a clue what actually turned into revenue. Then on targeting, they were throwing budget at low-income segments and age ranges that had zero buying intent. And here's the kicker: optimised targeting was enabled inside their retargeting campaigns. Which means the platform was expanding beyond the actual retargeting audience. More than $15K of that retargeting spend never even touched a qualified returning user.
Someone in the thread mentioned seeing audiences set to 'Observe' instead of 'Target' - classic. And someone else asked if that's wrong. Yeah, it's wrong. Observation means the platform can still optimise toward that audience, but you lose control. For retargeting, you want it set to Target only, or you're just burning cash.
After we fixed the tracking and targeting, conversion rates jumped 166%, we got over 1,500 trial signups and 900 installs, and the company attributed north of $1M ARR to the rebuild in a single quarter.
The lesson? Everyone wants to scale before they validate what the platform is actually optimising toward. Small measurement mistakes compound fast when spend goes up. The biggest gains don't come from new creatives or bigger budgets. They come from cleaning up data quality, tightening targeting, fixing attribution, and making sure the algorithm learns from the right signals. Otherwise, you're just feeding the machine garbage.
