A useful website SEO report connects an observation to its page, evidence, next action and follow-up result. It should tell you what was checked, when it was checked and what remains unknown. A crawl finding, a Google search click and a registration are different measurements.

This example includes a real public HTML snapshot and a 12-check report worksheet. Both downloads are free and require no account or email. The snapshot was prepared independently for this article; it is not an export from the AppStruct SEO product. No customer data or authenticated pages were collected.

What should a website SEO report contain?

PartRecordUseful next question
ScopeExact URLs, time, collection method and exclusionsWhich pages and behavior were actually checked?
FindingObserved value, affected URL and reproducible evidenceCan another person confirm the observation?
ActionProposed change, owner and a retestWhat would resolve this particular issue?
Search outcomeSearch Console period, page, query, clicks and impressionsDid Google show this page, and did someone click it?
Business outcomeWebsite source, engagement and the chosen conversion eventDid the visit lead to a useful product action?

Keep missing data visible. A tool's unavailable backlink or performance metric is not a measured zero. A report score also depends on the tool's checks and weighting; it does not establish a Google ranking.

Read the three-page HTML example

On 17 November 2026 at 00:08 UTC, we requested three public AppStruct pages. Each request fetched the initial HTTP response and parsed its HTML using Python's standard-library HTML parser. JavaScript was not executed.

PageHTTP responseTitle found in initial HTMLCanonical found
Homepage200AppStruct — App Builder & Collaborative Workspacehttps://appstruct.ai/
App Builder200No-Code App Builder for Mobile & Web Apps | AppStructhttps://appstruct.ai/app-builder
SEO product page200Website SEO Audit & Search Console Reports | AppStruct SEOhttps://appstruct.ai/seo

The JSON snapshot also records the final URL, H1 text, meta description, robots meta values and the X-Robots-Tag response header. At collection time, each page had a description and a robots meta value containing index, follow; no X-Robots-Tag header was returned for these responses.

These observations show that the sampled responses contained the listed metadata. They do not establish robots.txt permission, Google's selected canonical, current indexing, useful search positions, a complete site audit or a fast mobile experience. A future deployment can also change the values.

Reproduce one observation before proposing a fix

For a public page you are authorized to inspect, save the response headers and initial HTML:

curl --silent --show-error --dump-header headers.txt --output page.html https://appstruct.ai/seo

Inspect the HTTP status in headers.txt. In page.html, locate the title, description, canonical and robots declarations. This command does not follow a redirect or render JavaScript. If the response redirects, record that response and its destination before inspecting the intended final page. Keep private URLs and authenticated responses out of shared reports.

Compare initial HTML with the rendered page when the site relies on JavaScript. Use Search Console's inspection tools for Google's observations of a URL. Our PWA publishing checklist also separates public-page discoverability from installation and offline behavior.

Turn a finding into a bounded repair

Suppose a fictional page /collection has no title in the collected HTML. Record that exact observation, rather than describing the entire website as broken. Confirm that the page should be public, write a title describing its content, deploy the change and collect the same URL again.

Google's title-link guidance recommends descriptive, concise titles. Its snippet guidance explains that snippets may use page content or a meta description. Filling a field therefore does not force Google to display that exact text.

Handle canonical and indexing findings separately. A canonical declaration expresses a preference for duplicate or similar pages. A noindex rule controls indexing when a supporting crawler can read it. Neither is a replacement for deciding which content belongs in public search.

Keep technical checks and measured growth separate

Google documents that a successful HTTP response does not guarantee indexing. Similarly, resolving a missing title does not prove that a later traffic change was caused by that edit.

Keep a change date and compare complete periods with consistent filters. In Search Console, use the relevant pages and queries for impressions, clicks and average position. In website analytics, inspect the corresponding traffic sources and the actual event you defined as useful. Keep owner checks and QA traffic separate. Small samples need time before they support a reliable growth conclusion.

The 12-check CSV worksheet includes scope, response, metadata, rendered content, crawl controls, search outcomes and a retest. Every result starts as Not checked; the three-page example does not pre-complete the reader's audit. Add an owner and private evidence reference for each check.

Use AppStruct SEO for the next step

AppStruct SEO provides a workspace for reviewing technical findings with affected URLs, inspecting page metadata and connecting Search Console for measured search performance. Its public page includes a fictional interactive audit example; that example is separate from the real HTML snapshot in this article.

Using the product requires an AppStruct account and applicable access. Review the current options in your workspace; the free downloads here do not imply that every crawl or provider metric is free or available. Website changes still belong in your site editor or deployment workflow.

Start with one finding you can reproduce and retest. Explore the SEO workspace workflow, or use the launch tracker to keep public promotion links and their measured results in a separate worksheet.