Searching for an SEO audit report sample usually turns up PDFs designed to sell audits rather than demonstrate them — heavy on dials and traffic-light tables, light on anything you could hand a developer.
What follows is the part that actually matters: individual findings, written the way most reports write them, then rewritten. The structure they sit inside is covered separately in our SEO audit template. This page is about the writing.
What separates a finding from an observation
A finding has four parts. Drop any one and it stops being actionable:
| Part | The question it answers |
|---|---|
| Evidence | How do you know? Counts, URLs, dates |
| Consequence | What is it costing, in traffic or revenue |
| Fix | What specifically changes, and where |
| Effort | Hours, and whose hours |
Most audit findings supply the first part only, which is why most audits produce a list of things that are true and nothing that gets done.
Finding 1 — Indexation
As most reports write it:
⚠️ Indexation issues detected. Only 1,900 of your 4,000 pages are indexed by Google. Indexation is critical for SEO success. We recommend reviewing your indexation strategy.
Every sentence is true. The client cannot do anything with it. “Review your indexation strategy” is not a task anyone can be assigned.
Rewritten:
1,100 product pages have been deindexed since 12 June.
Evidence: Search Console shows indexed pages falling from 3,050 to 1,900 between 12 and 26 June. The 1,100 missing URLs are all
/product/pages. Spot-checking twelve of them, each carries<link rel="canonical" href="/category/...">pointing at its parent category. The category templates do not have this problem. The product template was changed on 11 June (commita4f21c8).Consequence: These pages were responsible for 41% of organic sessions in May. Current loss is roughly 12,000 sessions per month and will not recover on its own.
Fix: Remove the hardcoded canonical from
templates/product.htmlline 14 so it self-references. Resubmitsitemap-products.xml.Effort: 2 hours, engineering. Recovery expected within two crawl cycles of deployment.
Same finding. One is a warning; the other is a ticket.
Note what made the difference: a date, a commit reference, a file and line, and a number attached to the consequence. None of that required more skill — it required looking one layer further before writing.
Finding 2 — On-page
As most reports write it:
❌ 340 pages are missing meta descriptions. Meta descriptions influence click-through rate from search results. Best practice is 150–160 characters including the target keyword.
This is the single most common filler finding in the industry. It is technically correct and close to worthless — Google rewrites most meta descriptions anyway, and 340 hand-written descriptions is weeks of work for an effect nobody has isolated.
Rewritten:
340 category pages share one template flaw; the missing meta descriptions are a symptom, not the finding.
Evidence: All 340
/category/pages lack a meta description, an H1, and a unique intro paragraph. They render as a product grid with a breadcrumb. In Search Console these pages have a combined 42,000 impressions and 310 clicks — a 0.74% CTR against a 2.1% site average at comparable positions.Consequence: At site-average CTR these pages would earn roughly 880 clicks per month rather than 310. The gap is ~570 sessions/month.
Fix: Add an H1 and a two-sentence intro to the category template, both populated from the existing category name and description fields already in the CMS. This generates a usable meta description automatically. No hand-writing required.
Effort: 4 hours, CMS/template. Applies to all 340 pages at once.
The rewrite reframes a busywork task as a single template change. That reframing is the value of the audit. See meta descriptions and header tags for the underlying detail — but note that neither belongs in the audit body.
Finding 3 — Internal links
As most reports write it:
ℹ️ Internal linking could be improved. We noticed some pages have few internal links. Consider adding more internal links to improve link equity distribution.
“Could be improved” applies to every site ever audited. There is no evidence, no scale, and no place to start.
Rewritten:
31 of your 49 articles link to no commercial page.
Evidence: Crawled all 49 posts in
/blog/. Eighteen contain at least one link to a/services/URL; 31 contain none. Those 31 articles account for 61% of blog organic sessions.Consequence: The majority of your existing organic traffic has no path to a page that converts. This is not a ranking problem — the traffic already exists and arrives on pages with no next step.
Fix: Add one contextual link per article to the most relevant service page. A mapping table for all 31 is in Appendix C. Editorial placement in the body, not a footer block — sitewide blocks are discounted.
Effort: ~3 hours total, content team. No new content, no developer, no ranking dependency.
This is usually the best finding in any audit and almost never appears in one, because it produces no impressive-looking chart. It is also the only finding here that works this month regardless of what Google does. Our internal linking guide covers the method, and orphan pages covers the related case where nothing links to a page at all.
The pattern across all three
Each rewrite does the same four things:
- Names a number. 1,100 pages, 340 pages, 31 articles. Not “some” or “several”.
- States the cost. In sessions, not in severity ratings.
- Points at a file, a template or a list. Somewhere a person can go.
- Prices the work in hours. Which is what makes it schedulable.
If a finding cannot survive being written this way, it is probably not a finding. That is a useful test to run on your own audits before sending them: strip the adjectives and see what is left.
What a real audit looks like on delivery
- Sections 1 and 2 fit on two pages. Verdict, then a ranked action table.
- Supporting sections report only what is broken. No passing checks listed for volume.
- The appendix ships the raw data. Crawl export, URL lists, the Search Console query. If a supplier will not hand these over, the audit is a sales document.
An audit that cannot be summarised in one sentence has not finished its job. The auditor stopped at collection and left synthesis to the client.
Judging an audit you have been given
If someone has already delivered one and you are not sure whether it was worth paying for, run the four-part test above across ten findings at random. Fewer than half surviving is a reliable signal — and it lines up with the other patterns in signs of a bad SEO agency.
If the audit came from an agency you have since parted with, and you cannot reconstruct what was actually changed on the site, our cleanup service starts by rebuilding that history. If you simply need the audit done properly the first time, the traffic drop audit writes findings in exactly the format above — and what an SEO audit costs explains why the price varies as much as it does.
