The Google Ads reporting mistakes that lead to wrong decisions

← Back to Field Notes

I have sat in dozens of client reviews where decisions were made based on technically correct data that was being misread. Budget cuts to high-performing campaigns, continued investment in inefficient ones, and incorrect diagnosis of problems all stem from the same root cause: using the wrong metrics or the wrong frame of reference. Here are the most common examples.

Judging smart bidding campaigns on short time windows

Smart bidding strategies optimise over time. A campaign using Target CPA that has only been running for ten days is still in its learning phase. Evaluating that campaign's CPA against your long-term target after one week and concluding it is underperforming is a misread. The learning period for smart bidding is typically two to four weeks. Performance during learning is often worse than steady-state performance. Decisions about whether a bidding strategy is working should be based on performance after the learning phase, not during it.

Comparing date ranges with different seasonal profiles

This month versus last month comparisons are often meaningless if the months have different seasonal demand profiles. January versus December for a retail business will show apparently terrible January performance - but that is seasonal, not indicative of a problem. Period-over-period comparisons are only meaningful when compared against the same period last year, or when seasonal adjustment is explicitly accounted for. Always ask: would you expect this period to perform differently from the comparison period for seasonal reasons? If yes, the comparison needs context.

Blending brand and non-brand metrics

Brand campaigns convert at much higher rates and lower CPAs than non-brand campaigns. When you blend these together in top-line reporting, the overall account metrics look artificially positive. This matters particularly when evaluating whether non-brand acquisition campaigns are working. Separate reporting for brand and non-brand is not optional - it is essential for understanding whether your actual acquisition activity is efficient.

Treating all conversions as equal

If your account tracks multiple conversion types - form submissions, phone calls, page views, video views, live chat initiations - and counts all of them as conversions without differentiating by value or intent quality, your CPA figures are meaningless. A campaign that drives 50 page view conversions and 2 form submissions has a very different commercial outcome to one that drives 15 form submissions, even if the total conversion count is similar. Set your primary conversion actions to the actions that directly represent commercial intent, and review secondary actions separately.

Missing the view-through and assisted contribution

Last-click attribution, which remains common in Google Ads accounts, gives 100 percent of conversion credit to the final click. This consistently undervalues upper-funnel campaigns - YouTube, Display, Demand Gen - that influence conversions but rarely get the last click. If you are using last-click attribution and wondering why your brand awareness campaigns look terrible in the data, this is why. Use the Attribution report in Google Ads to see how conversion credit shifts when you move from last-click to data-driven attribution. The difference is usually illuminating.