Search Console: report non-branded clicks only, or keep brand queries in SEO totals?

Take a dental clinic whose organic clicks look flat month over month, but the number swings a lot depending on whether brand queries are included. Half the clicks are people typing the clinic name, and they'd have found it anyway. So one approach is to split Search Console with a regex on the brand name and common misspellings, then report only non-branded clicks as the SEO result.

The usual objection is that brand demand often comes from good content and reviews too, so stripping it out undersells the work. That's true, but it cuts both ways. If brand clicks jump because of a TV spot, crediting SEO for it makes the report less honest.

What I can't solve cleanly is misspellings. The regex always misses a few variants, and those land in the wrong bucket. How do you handle that?

Honestly, I'd stop trying to make the regex perfect. Misspellings are a small slice for most brands, and what matters more than which bucket they land in is splitting the same way over time. Brand volume changes that, though. If branded queries are most of your clicks, regex errors can swing the trend, so pull a month of queries and check what the regex misses before you trust the split.

Agreed on not chasing a perfect regex. I'd check first whether the split even adds up to your total. Search Console hides some queries for privacy, and those clicks count in the overall Performance totals but never show up in any query row, so branded plus non-branded will often come up short of the headline number. Filter by query, sum the rows, and compare that to the unfiltered total for the same period. The gap is your anonymized share. Report it as its own line instead of letting it fall into either bucket. That way the regex misses are bounded and visible, and the month-over-month trend stays readable even when the brand share moves.

The anonymized gap probably isn't neutral. Hidden queries are mostly rare, long-tail searches, and those tend to be non-branded, so the gap usually sits in the non-branded share rather than the branded one. If it swings a lot from month to month, your non-branded trend will be the first thing to move.

Also, Discover and News have no query dimension at all. Compare the query-level sum against the Web search tab only, otherwise the gap will look bigger than it really is.