We Scored Our Own Website First
Did you use your own tool on yourself?
Yes — and we'd argue you should distrust any AI-visibility tool that hasn't. If we're going to tell founders to make themselves legible to machines, our own site has to be the canonical example of it. So we score AIScoring with AIScoring, in public, and we're candid about what the number is and isn't. There's a name for this in software: dogfooding — using your own product in earnest. For a company whose entire job is measuring agent-readiness, it's not a cute gesture, it's the minimum credibility test. A scoring tool that can't or won't score itself is telling you something.
What did you score, and what does it mean?
On the deterministic transactability dimension — the part fully within our control — our target is a perfect 100. That means complete, valid schema; pricing exposed in machine-readable text; a clean, accurate llms.txt; and correct AI-crawler rules that let the agents reach what matters. This is the standard we hold you to, so we hold it to ourselves first, and we treat anything less than 100 on the parts we control as a bug in our own house. Why insist on 100 here specifically? Because it's deterministic and earnable purely through work. There's no excuse for a company that sells transactability advice to be less than perfectly transactable itself.
What about visibility — aren't you a brand-new domain?
We are, and we won't pretend otherwise — this is the honest part that matters most. LLM visibility is earned over time: it depends on being described consistently across sources the models have absorbed, and on day one a new domain simply hasn't earned that yet. Our visibility number starts low, exactly as it should for a site the world hasn't caught up to. Rather than hide that behind a flattering aggregate, we show it as "tracking from day one" — an honest, live demo of precisely what a new site looks like to the models and how the number should climb as we do the work we recommend. Our own visibility curve becomes part of the product story: watch us take our own medicine and see whether it works.
Why tell customers all this?
Because trust is the only durable moat a scoring company has. The moment we shade a number to look better, the entire product is worthless — a score you can't interrogate is just an opinion with a font. So we've made the defensible choices: our methodology is public and versioned, our own numbers are visible, our public score leads with the reproducible dimension, and we'd rather show you an honest starting line than a fiction that flatters us. It also keeps us honest operationally. Publishing our own score means we live under the same scrutiny we ask you to accept. If our advice doesn't move our own number, you'll see it, and so will we. Hold us to it — that's not a risk we're nervously accepting, it's the whole design.
Frequently Asked Questions
- Does AIScoring score its own website?
- Yes, publicly. We target a perfect 100 on the deterministic transactability dimension we control, and we show our visibility honestly as a new domain earning its reputation over time.
- Why is your visibility score low if your tool is good?
- Because visibility is earned over time and we're a new domain — the models haven't caught up to us yet. That's the expected starting point for any new site, and we show the climb rather than hide it.
- What is dogfooding?
- Dogfooding is using your own product in earnest, the way customers do. For a scoring company it's a basic credibility test: if we can't score ourselves well on the parts we control, why would you trust our advice?
- Can I see AIScoring's own methodology?
- Yes. Our methodology is public and versioned, so you can inspect exactly how the score is calculated and challenge any rule or weight you disagree with.
- Does a perfect transactability score mean you'll rank #1 everywhere?
- No — transactability and visibility are different. A perfect deterministic score means an agent can transact with us flawlessly; visibility is the separate, slower work of being widely and consistently described.