You might have noticed that in the Search Terms or Search Query Performance tab, when you add a filter to your view, the data changes versus when you don’t have a filter applied.
It might look like a bug, but it’s actually intentional.
When a filter is applied, the previous-period numbers only include the search terms + target combination that actually triggered for your filter in your currently selected period. If a search term matches your filter in the previous period but doesn't trigger again in your current period, it won't be counted where it doesn’t fit the filter criteria.
Let's look at an example.
Say a target was triggered by the search term "running shoes" and got 1 click in the previous period. If that exact target wasn't triggered by "running shoes" again in your current period, the combination drops out of the view, and that previous click isn't included in your previous-period number.
So the totals don't line up because you're not looking at every search term from the previous period.
You're looking only at the ones that match your filter, compared against themselves.
Because it's the more accurate way to compare. Search terms don't always get triggered by the same targets from one period to the next, so matching the exact combinations lets you compare apples to apples instead of lumping in terms that are no longer in play.
It's also a comparison you can't easily pull anywhere else. If you just wanted everything, the standard search term report plus some basic spreadsheet skills would give you that. But this view does the harder work of isolating the combinations that are still relevant, which is what makes it more useful.
No problem. Just remove any filters you've applied and change the date range to your comparison period directly. That gives you the complete totals without the combination matching.
