Writing · The audit

What does your website score actually measure?

Every website grader hands you a number. Ours shows the working: three scores, fixed weights, and the rules that stop the number lying to you.

Ricky Bell · 10 August 2026 · 4 min read

Shop owner counts coins into stacks on a wooden counter beside an open cash tin and a blank notebook.

Would you trust a consultant who wouldn’t show you their working? Neither would we. So before you run our audit — or anyone else’s — here is exactly how the number on the report gets made.

Three scores, one number

The audit produces one overall score out of 100, built from three inputs:

  • Speed — measured by Google’s PageSpeed Insights, on mobile, because the phone is where most of your customers actually meet your website. We take Google’s number as it comes; we never adjust it.
  • AI readiness — how much of the admin around your work is still manual. A contact form with no automated response, booking that only happens by phone, an FAQ answering the same ten questions by hand: each is a signal. Lower means more manual, which also means more opportunity.
  • Copy — whether a stranger can tell what you do, who it’s for, and why you, within about five seconds. Clear beats clever.

The blend is fixed: 30% speed, 35% AI readiness, 35% copy. Grades follow the score — 85 and above is an A, 70 a B, 55 a C, 40 a D, below that an F.

The odd thing, at first glance, is that the only externally measured input carries the smallest weight. That’s deliberate. The mobile lab test is the most objective number we have and also the least stable — the same site can move several points between runs. So speed informs the score without being allowed to dominate it.

The rules that keep the number honest

A score is only useful if you can trust what’s underneath it. So the audit runs under a small set of hard rules:

  • Nothing is scored from a placeholder. If a measurement fails, the report says “unavailable” — it never quietly substitutes an average and pretends it measured something.
  • Competitors are real. They’re discovered and verified before scoring, and the analysis is forbidden from inventing one to fill a slot.
  • No padding. If your site has two genuine issues, you get two — not eight thin ones to make the report look thorough.
  • Revenue estimates are labelled rough, and only offered when there’s something defensible behind them.

Why be this strict about a free tool? Because the audit is the first piece of work we ever do for you. If the free thing exaggerated to win your attention, you’d be right to wonder what the paid thing does.

What to do with your three scores

Read the three scores before the grade. Each points at a different kind of fix: speed is an engineering job, copy is a writing job, and AI readiness is usually a process job — taking repetitive admin off a human’s desk. The grade tells you how urgent things are; the three scores tell you where to start.

The audit is free, takes about a minute, and every number comes with its working attached. Run it, and argue with it if you like — it will show you why it said what it said.

Frequently asked

What is a good website audit score?

On the Fordi audit, 85 and above is an A, 70–84 a B, 55–69 a C, 40–54 a D, and below 40 an F. The grade matters less than the three scores underneath it, because each one points at a different kind of fix.

Why is page speed only 30% of the score?

Page speed is the only externally measured input, but mobile lab tests are also the most volatile — the same site can score differently run to run. So it informs the score without being allowed to dominate it, and the two judgement-based scores carry 35% each.

Does the audit compare me against real competitors?

Yes. Competitors are discovered and verified before any scoring happens — the analysis is forbidden from inventing one. Where only partial data exists for a competitor, they are scored on the real signal alone rather than blended with a placeholder.

← All writing