How Google Search counts results
A search recap comes from Google Ads. The results it shows are the ones Google credits to its ads, using whichever attribution model the account is set to. Search is the last step for a lot of buyers: they hear about you somewhere else, type your name, click the top result and buy. That puts search in a perfect spot to collect credit.
The most important split in any search report is brand versus non-brand. Brand search means ads shown to people who typed your name. Non-brand means ads shown for things like “hard seltzer near me” or “new horror movies.” Brand searches tend to show excellent results. Most of those people were already looking for you.
Search accounts also have settings that change the count: the conversion window, the attribution model, and which actions count as a result. Some accounts import results from Google Analytics as well as tracking them directly. Find the window in the report, and ask where each result comes from.
Where results get counted twice
- Brand search. Someone saw a TikTok, a trailer or a billboard, then searched your name. Search gets credit for the result. So does the channel that sent them.
- Mixed campaign types. Some Google campaign types run across search, YouTube, display and more at once. They can include searches for your name inside the mix. Ask for a split.
- Two sources for one result. If an account counts a result from its own tag and an imported one from Google Analytics, the same result can show up twice in one report.
- Email and direct. A buyer who clicks your email, leaves, then comes back through a search ad gets credited to search.
Last-click rules make all of this bigger, because they hand everything to the final step. Even accounts that use other models tend to reward search well.
A quick way to see the size of it: ask for one month of results split by brand and non-brand, then set the brand number next to what your other channels reported that month. If social, creators and search all report a big share, and the total runs past your own numbers, brand search is likely where much of the overlap sits.
A checklist for your next Google Search recap
| Look at | What to look for | Why it matters |
|---|---|---|
| Brand vs. non-brand | Results split by searches for your name | Brand search reaches people already on their way |
| Mixed campaigns | Whether brand searches sit inside broad campaign types | Brand credit can hide in the mix |
| Result sources | Which actions count, and where they’re tracked | Two sources can count one result twice |
| Model and window | Attribution model and conversion window | Both change how much credit search gets |
| Total vs. your numbers | Search results plus every other channel’s, against your total | Overlap shows up in the sum |
Questions worth asking
Ask them of whoever runs the account, in-house or a partner.
- How much of what search reported came from people searching our name?
- Do any campaigns include brand searches without labeling them?
- Which attribution model and window does the account use?
- Is any result counted from both the Google Ads tag and Google Analytics?
- What do we expect to happen if we pause brand search in one region?
That last question is a good one to ask out loud. Write down everyone’s guess before you run the test.
What a regions-on, regions-off test looks like on Google Search
Brand search is the cleanest place to start testing, because the question is simple: if we stopped paying for clicks on our own name, would those people still find us? Pause brand search in some regions, keep it on in matched regions, wait a few weeks, and compare sales in your own numbers. That’s a regions-on, regions-off test, also called a holdout test.
Two things to watch. Competitors may bid on your name while you’re paused, so note whether that happens. And keep everything else steady: no new promotions in one group only.
If results barely move in the paused regions, brand search was mostly collecting credit. If they drop, it was doing real work, and you have proof. Either way you’ve learned something worth more than the clicks. Our guide to incrementality without the math has a worked example.
Want to see what worked for your brand? Try it free: pick a question, and we answer it with your own numbers and what people say online. No card.