A first pass, not a verdict
Paste a public URL and the checker reads the page the way a crawler would. It reports on what it finds, then gives you a list of things to fix, ordered by how much they matter.
It doesn't use AI to judge your writing. It checks facts: is the page indexable, is the title there, are headings in order, do images have alt text. Those are things a tool can verify.
What it checks
- Indexability: status code, robots meta, canonical
- Metadata: title and description length and presence
- Headings: one H1, a sensible outline
- Links: internal and external counts, broken anchors
- Images: alt text, size hints
- Schema: JSON-LD found and parsed
- Social: Open Graph and Twitter tags
- Answer engine signals: clear questions, direct answers, structure
AEO and GEO signals
Search is moving toward answers. Pages that state a clear question and answer it near the top, use lists and tables where fitting, and carry schema are easier for answer engines to quote. The checker looks for those traits.
It can't promise you'll be cited. It can show you what's missing.
A high score isn't a ranking
A page can pass every check and still lose to one that answers the question better. The tool can't tell if your content is good. Read the findings with your own judgment, then ask the harder question: would I want to read this?
How to use it
Run it on pages you edited, after each release. Fix errors first, warnings next. If a page is blocked from public access, it can't be checked. Nothing you type is stored.
Reading findings in order
Not every warning matters equally. Fix blockers first: a noindex tag you didn't mean to add, a canonical pointing somewhere odd, a page that returns an error to bots. Those can remove a page from search entirely.
Then look at titles and descriptions, headings, image alt text and internal links. These are the everyday improvements that help clarity for both people and search engines.
Last come the extras: structured data, social tags and signals for answer engines. They help, but only after the basics are fine.
A sensible workflow
Run the checker on one page, make a handful of changes, then run it again. Don't chase a perfect score on every page. Spend your time on the pages that bring in or could bring in visitors.
Compare against a competitor's page too. Seeing what a top result includes that yours doesn't is often more useful than the score itself.
Keep a short log of what you changed and the date, so you can link movements in traffic to your edits.
The checker fetches the public page you enter. It can't see content behind logins, and it doesn't replace reading your pages the way a visitor would.
- Check indexability first
- Fix titles and descriptions next
- Review headings and links
- Add structured data where it matches the page
- Re-test after each round of edits