How to Turn a One-Off Audit Into a Monthly Retainer

Published 2026-09-20 · 9 min read · by the isaeo team

Why most audits end at one invoice

You send a thorough 40-page PDF. The client says thanks. Nothing happens. Six months later they hire someone else to write another one.

The failure is structural, not a quality problem. A finished document implies finished work. If the deliverable reads as a verdict — here is what's wrong, good luck — then the natural next step is silence. Nothing in the artifact suggests a month two.

Deliver a baseline, not a verdict

Change one thing: make the audit obviously the first of a series. Three elements do that work.

• A number. Not a grade — a score with a stated method, so it can go up. "Health 64, AEO 31" invites "what will it be next month?" in a way that "several critical issues found" does not.
• A split fix list. Every item assigned to someone: you, their developer, or their content person. A list with no owners is a list nobody starts.
• A booked re-crawl. "We'll re-run this on the 14th" turns the audit into the first measurement rather than the only one.

None of that requires more work than you were already doing. It requires the work to be shaped as an ongoing measurement.

The retainer structure that works

By month three the value has quietly shifted from "finds problems" to "notices regressions", which is a much stickier reason to keep paying. Their developers ship every week; something breaks eventually; you're the reason it gets caught in days instead of months.

Month two is the real pitch

Month one sells on findings. Month two sells on movement — and movement is only visible if you kept the before. This is the mechanical reason change tracking matters commercially and not just technically: the diff is the sales asset.

Put it at the top of the second report. Three lines:

• Health score 64 → 79.
• Closed: 14 issues. Opened: 2 (both from last Thursday's release).
• Still blocked on your side: 3 items, listed below.

That last line is doing a lot of work. It makes the client's own delivery visible in your report, which is both fair and quietly effective at getting the remaining items done.

How to price it

Two defensible models, and one to avoid.

1. Audit paid, retainer discounted. Charge full price for the audit, credit it against the first month if they sign. Clean, and it prices the audit honestly rather than pretending it's free.
2. Retainer only, audit included. No separate audit fee, three-month minimum. Simpler to sell, worse for tyre-kickers.
3. Avoid: free audit, hope for retainer. A free deep audit is unpriced work that attracts people who want a free deep audit. Use a short, automated free audit as the pitch instead, and keep the deep one paid.

On the tooling side, the maths only works if adding client eleven doesn't add eleven hours. That's the per-site cost you should be checking on any plan you consider.

Three mistakes

Sending everything. A 1,400-issue export reads as unfixable, and unfixable reads as not worth paying for. Send the top ten with severity and effort. See reports clients understand.

Reporting activity instead of outcome. "We audited 340 pages" is your effort. "Eleven pages are back in the index" is their outcome. Only one of those survives a budget review.

No regression story. If you never show them something you caught, the monitoring value stays invisible and the retainer looks like a subscription to a document. Name the catch explicitly when it happens.