Type three letters and the box offers full phrases. Those phrases come from what many people have searched, filtered by language and by country, and sometimes nudged by a signed-in history. A suggestion that appears in France can be absent in Canada. A suggestion in English can be absent in the same country for another language.
That is the first practical point. A placement is a country-and-language object. “We want to be global” is not a brief. “We sell in the United States, in English” is.
Volume is not the whole story
Popular queries are more likely to have a stable list, because lots of people finish them in similar ways. A rare phrase may have no suggestions at all, or a list that changes every week. Chasing a phrase nobody types produces a screenshot of an empty box.
The other ingredient is association. A brand name shows up beside a category when the two have been seen together often enough, in public, that the system treats them as a pair. A consistent name, a page that actually covers the category, and mentions that use the same spelling all help that pair exist. A logo buried in a PDF does not.
What you can measure
Measure the thing you can see. The exact query, the country, the date, and a clean browser that is not signed into a personal Google account. Personalized suggestions are a different list. If you only check while logged in, you are grading your own history.
- Write the query character for character, including plurals.
- Check the country you paid for, not the country you happen to be in.
- Keep the screenshot with the date. A suggestion can leave as well as arrive.
What you should not expect
A permanent rank. Autocomplete is recomputed. A name can enter the list and later drop if the association weakens or if many other completions become more likely. Treat a placement as something you watch, not a trophy you file.
You should also not expect the list to override a mismatch. If the company does not offer the thing the query names, public pages will not support the pair, and the suggestion is the wrong goal anyway.
Scalesuggest works inside those limits. We pick a query with demand, in a country you sell to, and we build the public association. We do not sell a position Google does not offer.