I audited my own SEO properly for the first time. I'd been wrong about three things.
Hi all!
I've been doing SEO on my SaaS for months on instinct. This week I finally pulled real data. Search rankings, competitor link profiles, keyword difficulty and checked my assumptions. Three of them were wrong, and the pattern is embarrassing enough to be worth sharing.
- I was targeting the harder keyword on almost every tool page.
My font tool page was titled "Pinterest Font Generator". The obvious keyword, right? Except:
"pinterest font generator": 260 searches/month, ranking pages average 26 referring domains
"pinterest fonts": 1,000 searches/month, ranking pages average 3 referring domains
Four times the volume at a ninth of the competition, and I'd picked the wrong one because "generator" sounded more like what my page was. I had this backwards on several pages.
- Keyword difficulty scores are close to useless on their own.
I found a keyword at difficulty 11 with 1,000 searches and got excited. Then I checked how many referring domains the ranking pages actually had: 48. Another at difficulty 17: 226 domains.
Worse, "referring domains" alone can also lie. I found a batch of keywords with one referring domain and huge volume, they were all on domains like Shopify and Pinterest, where domain authority does the ranking and individual pages need no links at all.
The only check that consistently worked: open the actual search results and look at who's there. Everything else is a proxy that fails in both directions.
- My competitor comparison content was completely dead.
I'd written a pile of "us vs them" and "alternative to X" articles. Combined impressions across most of them: zero. Not low, zero, never.
Obvious in hindsight. "MyProduct vs Competitor" requires people to already know MyProduct. Comparison content is a post-traction play and I wrote it pre-traction. I'm leaving the pages up because LLMs seem to cite that format when people ask them to compare tools, which may end up being the only reason it was worth writing.
The thing I keep relearning: every one of these mistakes was me trusting a number that was easier to get than the right one. Difficulty score instead of the SERP. Volume instead of intent. What the page is instead of what people type.
If you're doing your own SEO, the cheapest habit I can recommend is: before you write or retitle anything, open the search results for the exact phrase and look at who ranks. It takes thirty seconds and it would have saved me months.
The pivot from high-volume, high-difficulty keywords to those that actually align with your site's current authority is the most undervalued step in SEO. We often get caught up in vanity metrics or 'obvious' search terms, but the reality is that chasing a keyword with a 26-domain-backlink barrier when you are starting out is just burning cycles. Focusing on lower-competition, long-tail terms might feel like you're aiming smaller, but it is the only way to build the steady traffic floor you need to eventually rank for those head terms.
The comparison content pre-traction mistake is one I watch founders repeat. I did it too. That keyword requires people to already know you exist. The SERP tells you in 30 seconds: if every result is a third-party review site, you're not getting in without the brand recognition to match. The insight on difficulty vs referring domains is the important one. A difficulty 14 keyword I chased turned out to be a SERP full of Shopify subdomain pages. Difficulty scores break in both directions depending on who's actually ranking. Have you found a tool that does the SERP check automatically, or is it still manual?
Finding three assumptions were wrong once you looked at the underlying evidence is the interesting part.
Has the audit changed just what you’re doing in SEO, or has it changed how you decide which growth assumptions are trustworthy enough to act on in the first place?
The "vs competitor" point hit hardest — did the same pre-traction, got the same zero. The part that clicked: comparison pages don't just need volume, they need people to already know the frame of reference. "MyProduct vs X" requires brand awareness; "tool that does Y" doesn't. One is post-traction SEO, the other is discovery SEO. We now build every page around the discovery phrasing and treat comparison content as a later asset, not a launch play.