How to present a website audit to a business owner, so they act on it

Most website audits fail after they are finished. The measurements are right, the tool ran properly, and the business owner opens a 40-page PDF, sees a grade and a wall of amber, and closes it. Nothing happens, not because the problems are not real but because the document was written for the person who made it.

A business owner is not asking "what is wrong with my website?" They are asking "is this costing me money, how much, and what do you want me to do?" Here is how to build an audit that answers that.

Lead with three findings, not forty

The first page should hold the three things worth fixing first, chosen for the service you actually sell. A web studio leads with conversion and speed. An SEO consultant leads with indexability or AI visibility. A security provider leads with exposure and email spoofing.

Everything else still belongs in the report, further down, for the developer who will do the work. But the owner's decision is made on page one, and three is roughly how many problems a person can hold in their head while deciding.

If you need a way to choose, rank by three questions: How severe is it? How confident are we that we measured it rather than inferred it? Can we fix it and prove we fixed it? A finding that scores well on all three is a finding you can sell.

Translate every finding into a consequence

Compare two versions of the same finding:

Largest Contentful Paint: 5.4s (poor).

On a phone, your home page takes 5.4 seconds to show its main content. Google's threshold for a good experience is 2.5 seconds, and most visitors on mobile do not wait that long before going back to the search results.

The second one says what was measured, against what standard, and why it matters, in a sentence the owner can repeat to their business partner. That is the whole trick, repeated for every finding on page one.

For the technical reader, keep the raw measurement and the method beside it. The owner and the developer are reading the same document for different reasons, and it should serve both.

Put a number on it, and show your working

Owners respond to money more than to grades. Where you can, estimate what a problem plausibly costs: enquiries lost to a slow mobile page, invoices landing in spam because of a missing DMARC record, customers who cannot find the business in the map results four streets away.

The estimate must come with its assumptions, stated plainly. "Based on your listing's review volume, a typical conversion rate for your trade and the mobile abandonment rate for pages over four seconds" is defensible. A bare figure is not, and the first owner who asks where it came from will stop trusting everything else in the report.

Separate measured from inferred

Some findings are measured directly: the certificate expires on a specific date, the DMARC record does not exist. Others are inferred: the site probably runs an outdated plugin, based on the files it serves. Label which is which. It feels like weakening your case, but it does the opposite. When the owner's nephew who "does websites" challenges one finding, you can show that you already said it was an inference, and the rest of the report survives.

The same applies to checks that did not run. If the accessibility scan timed out, the report should say so rather than show a clean score.

Give them the fix, not just the problem

A finding with no remedy is a complaint. For each one on page one, say what the fix is, roughly how long it takes, and who does it. Paste-ready fixes help here: the exact DNS record for their mail provider, the redirect rule, the schema snippet. Even if the owner will never paste anything, seeing a concrete fix makes the problem feel solvable, and solvable problems get budget.

Make it readable on a phone

Owners open the report from your email, often on a phone, often between jobs. A shareable web link that opens without a login works better than an attachment. Keep a PDF for the people who want to forward a file. Make sure both say the same thing.

Structure that works

A simple structure for a client-facing audit:

1. Summary: an overall grade, the three findings that matter most and an estimated impact, on one page.

2. Scorecards: one line per category with its score and what fed it.

3. Findings: each with what it means, why it matters, what it costs, how to fix it and a confidence label.

4. Action plan: the ranked work, with effort estimates you can price.

5. Evidence: screenshots and the raw measurements, for the developer.

We describe what goes into each section of a SiteAssay report in the reports documentation.

Deliver it in a conversation, not an inbox

The best audits are walked through, not sent. Book fifteen minutes, share the link, and go through page one together. Stop after each finding and ask whether it matches what they have noticed: fewer calls from the website, emails going missing, customers saying they could not find them. When the owner connects a finding to something they have already felt, the audit has done its job.

Then end with a single, specific next step. Not "let me know if you have questions", but "I can fix the email records and the mobile speed this month for a fixed price. Shall I send a proposal?"

Follow up with a re-audit

The most persuasive page you will ever show a client is the same measurement after the work is done. Re-run the audit once the fixes are live and show the before and after side by side. It proves the value of what they paid for and it is the natural start of a retainer.

If you are building your audit process from scratch, start with the website audit checklist. SiteAssay produces reports in this structure, with confidence labels, cost estimates and paste-ready fixes on every finding; see the features or the pricing.