Google Business Profile primary category: broad one or the money service?

Picking the primary category on a Google Business Profile looks like a minor form field, but it changes more than most people expect. Say a dental clinic does general dentistry, cosmetic work and sleep apnea screening. If you choose 'Dentist' as the primary, you keep the broad local searches, but the cosmetic side may struggle to show up for 'cosmetic dentist near me', because Google leans on the primary category before it reads the services list. Switching to 'Cosmetic dentist' can narrow the listing's reach and cut the general terms that bring in routine bookings. My view is to pick the category that matches the service with the best margin per call, then use the services section and posts to cover the rest. Has anyone split this out with call tracking on, or is it mostly guesswork?

I'd push back a little on choosing by margin alone. The primary category acts as the filter for which searches you show up in at all, while the services list is a much weaker signal. So a "Cosmetic dentist" primary could cut the routine bookings you depend on, and you'd barely notice, because you only see the calls you still get.

My rule of thumb is to pick the primary that matches the service you want more of over the next year, as long as it still describes the whole clinic honestly. The services list, posts and photos can cover the secondary work. If two categories are close, go with the broader one.

On tracking, I don't think you can split this cleanly. GBP calls don't carry category data, so the only real test is changing the primary and watching call volume for a few weeks, ideally with a separate number on the profile. Seasonality makes that messy, so I'd treat the result as directional.

You don't have to rely only on the calls you still get. The Business Profile performance report lists the search queries that surfaced your listing, so you can compare before and after the primary category change and see whether routine terms like "dentist near me" drop off, whether or not anyone calls. The data is thresholded and incomplete, so low-volume queries won't show up. Even so, it answers the reach question directly, where call volume is a lagging and small signal for a clinic.

The query report only gives impression ranges, not exact counts, and it's grouped by month. So compare full months and leave the switch month out of both baselines. A category change can take a few weeks to settle, which means checking on day 10 might show a dip that reverses later. Pull a baseline of at least two or three months before the change, then compare again after a few weeks.

Agreed on the ranges, and I'd lean on ratios more than raw counts. Each month comes back as a bucket, so a move from 50 to 150 impressions can look like nothing or like a big jump depending on where the bucket edges fall. Tracking routine terms as a share of all queries in the same month cancels out a lot of the seasonal noise, since checkup season and the year-end insurance rush hit both sides. The calendar matters too. If the switch happens in March, compare against the previous March if you have it, because a three-month baseline from a different season can mislead you. Once the query share is moving the way you expect, call numbers on a tracking line make a handy sanity check.

Ratios only help if the terms you care about clear the reporting threshold in the first place. For a single-location clinic, "dentist near me" will usually show up every month, but something like "sleep apnea dentist" or "cosmetic dentist [town]" might sit below the line, and then your share calculation is mostly noise. So it depends on query volume. Look at which terms you'd actually lose before you pick a metric, and if they're thin, lean on the tracking line and accept a slower read.

Depends on how much of your revenue comes from those thin terms. If sleep apnea bookings are a small slice, the tracking line works fine as a proxy. If they're where your margin lives, a slow read from calls alone won't tell you much, so you'd want a longer test window.