Open Ahrefs, Semrush, or your tool of choice. There’s a "topical authority" score. It’s pretty. It correlates with ranking. Every agency’s monthly deck has it.
We’ve never used it to make a real call.
Here’s why the metric is mostly vibes — and the three numbers we actually look at instead.
What topical authority claims to measure
Different tools calculate it differently, but the pitch is roughly: "a score from 0–100 representing how authoritative your site is in a given topic cluster, based on the breadth of related content you publish, the backlinks pointing to that content, and the engagement on those pages."
The score is supposed to be predictive of how well you’ll rank for new keywords in that cluster.
Why it’s mostly vibes
Two problems with the metric.
First, it’s circular. The "authority" score is calculated from signals that already tell you whether you rank — backlinks, content depth, engagement. Saying "your topical authority is 73 so you’ll rank for new keywords in this cluster" is roughly equivalent to saying "your existing rankings predict your future rankings." Which is true. But it’s not insight, it’s tautology dressed up in chart form.
Second, it averages the wrong thing. The score smooths across an entire cluster. But Google doesn’t rank you on cluster-level signals; it ranks you on page-level signals modulated by site-level signals. A site can have terrible "topical authority" by these scores and still rank #1 for the highest-value query in the cluster, because that one page has the right backlinks and the right engagement profile.
We’ve seen this in our own data, repeatedly:
- Brands with topical authority scores in the high 70s ranking #15 for their flagship category keyword.
- Brands with topical authority scores in the 30s ranking #2 for the same category, because they have one excellent page with great links.
- Brands whose topical authority score moves +20 over six months without their actual revenue moving at all.
The score isn’t useless. It’s just not actionable. You can’t do anything with it that you couldn’t do better by looking at the underlying signals directly.
What we actually track
Three numbers, every month, every client.
1. Top-10 keyword count by commercial intent
Not total keywords ranked — total keywords ranked in the top 10 for queries we’ve classified as buyer-intent. This filters out the "we rank for 12,000 keywords!" noise where 11,800 are zero-volume long-tail garbage.
2. Click-through rate on ranking keywords
If you’re ranking #4 for 200 buyer-intent keywords but your CTR is 1.2%, you have a meta description and title problem worth more than another month of link building. CTR is the cheapest fix and most agencies skip it.
3. Branded search volume month-over-month
This one is the killer. Branded search lift is the cleanest indicator of whether your top-of-funnel work — AI citations, PR, listicles, podcast appearances — is actually building demand. It’s downstream of everything good a real brand SEO program does. We’ve seen it lift 200%+ in 6 months on programs where "topical authority" barely moved.
The right way to use SEO tool scores
Use them for diagnostic snapshots, not for tracking. A topical authority score gap between you and a competitor tells you something useful about where to invest. The same score tracked month-over-month for your own site is mostly noise — and worse, it’s noise that lets you congratulate yourself for activity that isn’t producing revenue.
We bet our retainer on top-10 buyer-intent keywords, CTR on ranking pages, and branded search lift. The rest is decoration.