Our page ranks #5 and gets zero clicks. Took me way too long to work out why.

We run a small astrology wiki on the side. Last month one of its pages was sitting at position 5 for its main query, 142 impressions in a week, zero clicks.

My first assumption was that the title was bad, so we rewrote a batch of titles and meta descriptions. Nothing moved at all.

Eventually I gave up on theories and just searched all nine of our zero-click queries by hand. Six of them had an AI Overview at the top answering the question completely. Birth date, sign, traits, all of it. There was nothing left to click for.

The other three had no AI Overview. Those results were full of tools that generate a chart on the spot, and we were turning up with an article. Same outcome, different reason.

Then I split the whole site by rank band:

position 1-6: 469 impressions, 26 clicks (5.5%)

position 7-10: 3,206 impressions, 17 clicks (0.53%)

Around 70% of our impressions sit in that second bucket. I'd been reading that as "we have traffic". We don't really. We have a number that looks like traffic.

The part that annoyed me most came after. I spent two days calling this a disaster because I was measuring against a 2% CTR benchmark I had in my head, and that number is from before AI Overviews were everywhere. The current figure for that position with an AIO present is somewhere around 0.65%. So we're slightly under par in a bucket that's bad for everyone, not uniquely broken. Our top-6 pages are actually above benchmark, which I never checked because I was busy staring at the bad number.

That's more or less why we're building GenGrowth. Every SEO tool I've used is good at telling you what's wrong on a page. None of them tell you whether fixing it is worth the week it costs you. We kept pouring time into things that were never going to rank.

Four tools are live and free. Two don't need an account, you just paste a URL and get a site-wide audit or an internal link map.

Where it falls short, and the thing I actually want feedback on: it's decent at observation and weak at judgement. It'll cheerfully tell you "66 of 162 pages have a different canonical target", which is accurate and completely useless if you've never touched a canonical tag. I know the gap is there. I don't have a good sense of how much it matters in practice.

If you try one, I'd like to hear the exact moment you thought "ok... and now what". That's more useful to me than someone saying it looks nice.