In this guide
Key takeaways
- A good report frames the problem, proves each finding with evidence, links it to a business outcome, ranks it by impact, effort and risk, and assigns an owner, a date and a way to verify the fix.
- Raw tool exports are supporting material, not the report. Google notes that third-party tool scores are the vendor's own, not Google's.
- Google says no one can guarantee a number one ranking and advises finding another SEO if someone does. A guarantee in a report is a reason to stop reading.
- A report passes my Monday-morning test if a developer or content writer could start work from it without a call.
What a good report is for
After years of writing and reading audit reports, I judge them by one thing: do they make decisions easy? The good ones do. The weak ones are long, look thorough and leave the reader unable to say what to do first.
An audit is a diagnosis. It cannot make rankings move by itself, and it should never pretend to. Its value is in telling you what is limiting visibility, how sure the auditor is, and in what order to fix things. If you are still choosing which type of audit to buy, Technical, Complete or Content: Which Audit Do You Need? will help you pick one first.
The anatomy of a good report
| Section | What it should contain | Why it matters |
|---|---|---|
| Summary | One or two pages: what is wrong, what it costs the business, the top five actions | Decision makers can act without reading everything |
| Scope and method | What was checked and what was not, the data sources, the date of extraction and any limits such as crawl size | You can judge how far to trust the findings |
| Findings with evidence | Each issue with example URLs, counts, screenshots or data, and how to reproduce it | Anyone can verify it, including your developer |
| Priority | A rating by impact, effort and risk, with the reasoning | The first month's work is clear |
| Action plan | Task, owner, due date and the check that confirms the fix worked | Findings become a project that can be tracked |
| Appendix | Full exports and lists | Supporting detail, kept out of the way |
Problem framing
Every finding should be stated in terms of its effect: "product pages cannot be indexed" rather than "meta robots directive detected". The business consequence comes first and the technical detail second.
Evidence and data provenance
You should be able to see which tool or report each number came from and when it was captured. Google's own data from Search Console carries more weight for questions about Google, and Google points site owners to it for that reason. A crawl export is useful for finding candidates, but it reflects one crawler's view on one day.
Business relevance
The report should connect issues to the pages and goals that matter: the pages that earn leads, the markets you serve, the products you want to sell. It may estimate the direction of an effect, but it should not promise an outcome.
Owners, dates and validation
This is the part most reports leave out. Each task needs a named owner, a realistic date and a check, such as "the URL returns a 200 in the live test and the page shows as indexed within two weeks". Without it, an audit is a list of opinions. On large sites, SEO governance is what makes those owners and dates stick.
One finding done well
Here is an illustrative finding, written the way I would present it. It is hypothetical and not drawn from a real client.
| Finding | Product category pages carry a noindex directive after a template update |
|---|---|
| Evidence | The meta robots tag shows noindex on all 14 category pages tested. The Page indexing report lists these URLs as "Excluded by 'noindex' tag". Example: three category URLs with screenshots of the tag. Data captured on a stated date |
| Why it matters | The categories are the main landing pages for commercial searches, so none of them can appear in results |
| Recommended fix | Remove the directive from the category template and re-deploy. Confirm that the sitemap still lists the URLs |
| Priority | Critical: high impact, low effort, low risk |
| Owner and date | Developer lead, within the next release |
| How to verify | Live URL test shows the page as indexable; the Page indexing report moves the URLs to indexed over the following weeks |
Notice what is missing: no promise of a ranking, no vague "improve your technical SEO". Everything can be checked.
What a complete report covers
- Crawling and indexing: robots.txt, sitemaps, canonicals, redirects, status codes, noindex rules and rendering. See Why Your Important Pages Aren't Getting Indexed for how I diagnose these.
- Page experience: Core Web Vitals on the main templates (Google's targets are LCP within 2.5 seconds, INP under 200 milliseconds and CLS below 0.1), HTTPS and mobile usability.
- On-page and content: whether pages answer the search intent, cover the topic and are not duplicated. Google's starter guide says a good title is unique, clear and concise, and a good meta description is short and unique, so I check those qualities rather than apply a formula.
- Internal links: orphaned pages, click depth and anchor text that tells readers what they will find.
- Links from other sites: relevance and risk, not a count. Third-party authority scores are the vendor's own measure, and Google says third-party tools do not have access to its ranking data. Google also says most sites do not need the disavow tool, which is meant for sites with a manual action or a real link scheme.
- Structured data: valid markup that matches the visible page, and used only where it enables something worthwhile.
- Measurement: whether GA4 and Search Console are set up to show what the business needs to know, such as key events from organic traffic.
Where relevant, a good report adds local, international or ecommerce checks. A report that cannot tell you whether the crawl covered your whole site is not finished.
How priorities should be set
I weigh each finding on three things. Impact is how much visibility or revenue is affected, judged by the pages involved. Effort is the time and the people needed. Risk is the chance that the fix breaks something else. Critical findings block crawling or indexing of important pages. High findings waste or split signals on many pages. Medium findings weaken pages that are already indexed. Low findings are cosmetic. The scale is simple on purpose, so that a non-technical manager can follow the reasoning. A longer explanation of this triage, with examples, is in the 12 audit errors that actually matter.
Estimates of improvement should be ranges with stated assumptions, or just a direction. Google's guidance for hiring an SEO says to ask what results to expect and on what timeframe, and that an audit should give realistic estimates, not promises.
Red flags in a bad report
- A raw tool export with a cover page. If you can recreate the document by running a crawler, you have paid for a printout.
- A single "SEO score" as the headline. Scores are the tool vendor's own opinion, not Google's.
- Guaranteed rankings or traffic. Google states that no one can guarantee a number one ranking and says to find someone else if an SEO guarantees first place. It also warns against anyone claiming a special relationship with Google or selling a "priority submit".
- Vague advice. "Improve content quality" or "build more links" with no pages, no scope and no owner.
- Generic text that could describe any site. A real report names your URLs and your templates.
- Recommendations that break Google's guidelines. Buying links, mass-submitting to directories, hidden text or stuffing keywords.
- Mass disavowing as a routine fix. Google says most sites do not need it, and misuse can harm performance.
- No limits stated. Nothing about what was not checked, how big the crawl was or when the data was taken.
- Secrecy. Google advises being careful about anyone who will not clearly explain what they intend to do.
A 15-minute test of any report
- Read only the summary. Can you say what the top three problems are and why they matter to your business?
- Pick one finding and verify it. Open the example URL. Does the evidence match what you see?
- Look for an owner and a date on at least the top ten tasks.
- Check the limits. Does it say what was crawled, which data was used and when?
- Apply the Monday-morning test. Could a developer or writer start the first task without a call?
- Scan for guarantees. Any promise of a ranking or traffic number fails the report.
Before you commission an audit
Google's page on hiring an SEO is the best short checklist I know, and it is worth reading in full. The points I would underline:
- Ask for examples of past work and how success was measured.
- Ask how they will communicate, and whether they will share every change they make and the reasoning behind it.
- For an audit, give read-only access to Search Console at first and do not grant write access.
- Remember that you are responsible for the actions of any company you hire.
Scope and price go together. For typical market ranges in Nepal and how to compare quotes line by line, read Cost of SEO in Nepal. I do not publish a rate card, because the right scope depends on the site.
How I structure my own audits
I build every audit around five questions. Can search engines and people find the important pages? Can they understand each page? Is each page relevant, useful and trustworthy enough to compete for the search? Can visitors take the next step? And which changes will create the most business value? The last question decides the order of the report. You can see the formats I offer, from technical and on-page audits to migration audits, on the SEO audit service page. If your audit is tied to a planned redesign, the pre-launch version is described in the SEO migration audit checklist.
An audit is the beginning of the work, not the end. After the fixes go live, re-check the same evidence. Google suggests waiting a few weeks before judging a change, and says some changes take months. My own reports include the check for each fix so the follow-up is mechanical.
Frequently asked questions
How long should an SEO audit report be?
Length is not a quality measure. A short summary with a prioritised action plan, followed by findings and evidence, is better than a long document with no order. The appendix can be as long as it needs to be, because nobody has to read it end to end.
Is a tool-generated audit enough?
It is a useful first pass. It finds candidates, but it cannot tell you which ones matter for your business, and its scores are the vendor's own. Interpretation and prioritisation are what you are paying for in a human audit.
Should a report estimate how much traffic I will gain?
It can describe the likely direction and the assumptions behind it. It should not promise a number. Results depend on competition, content, links and what else changes on the site.
How often should I get an audit?
After any major change such as a redesign or replatform, and when something unexplained happens to traffic or indexing. In between, regular checks of the Page indexing and Crawl stats reports catch problems early.
Who should receive the report?
The decision maker should get the summary, and the people doing the work should get the findings and tasks. A report that serves only one of those audiences usually fails the other. For the decision-maker layer, How to Report Enterprise SEO to Executives shows the one-page structure I use.
Can I do an audit myself?
You can do a good first pass with Search Console and free tools such as the Website Health Check. If you want to learn the full process, the Advanced SEO course has you produce a complete audit report as a portfolio piece, and SEO Jobs in Nepal explains which roles use that skill.
Sources and further reading
Technical statements about Google in this guide are based on the pages below, checked on 17 October 2026.
- Do You Need an SEO? Tips for Hiring an SEO: red flags, read-only access and what to ask
- Google Search's guidance on third-party SEO tools and advice: tool scores are not Google data
- SEO Starter Guide: titles, descriptions and how long changes take
- Disavow links: when the tool is and is not appropriate
- Page indexing report: statuses used in indexing findings
- Core Web Vitals: LCP, INP and CLS targets
- Google Search spam policies: link spam, hidden text and keyword stuffing