Sample AI SEO audit

SEO for AI Agents

flowforge.example.com

AI visibility not measured yet.

We have not checked AI engine answers for this site yet. The AI Visibility score below measures how ready your pages are for AI engines, not whether they recommend you.

AI Visibility52/1007 findings across three pillars
B2B SaaS product siteSample
Homepage readiness score: 64 out of 100. Readiness 64, outcomes not measured yet.

Readiness 64, outcomes not measured yet

1 page crawled

Measures how ready this page is for Google, Bing and AI answer engines. Does not measure rankings, traffic, whether AI engines cite you or real-user page speed.

Not measured: AI engine answers, Real-user page speed (field data)

AI Visibility score: 52 out of 100.
Technical score: 64 out of 100.
Foundational score: 76 out of 100.

No checks apply to this site

Local score: not applicable.
Run this on your site

This is a representative report for a saas site, built from illustrative data so you can see the full value before running your own audit. A real audit produces the same shape from your live pages, with a confidence flag and methodology link on every finding.

Synthesis

Executive summary

  • FlowForge scores 64 overall, the mean of Foundational, Technical and AI Visibility.
  • Local SEO does not apply, because the site shows no physical location.
  • Foundational is solid, but AI Visibility and Technical drag the score down, and the cause is the same: the marketing site is a client-rendered single-page app.
  • AI crawlers read an almost empty shell, so ChatGPT and Perplexity cannot cite the product even though the content is excellent in a browser.
  • The single highest-leverage fix is server-rendering the homepage, the features pages, and the pricing page.
  • After that, ship Organization and SoftwareApplication schema with sameAs links, restructure the docs and FAQ into extractable passages, and pursue a Wikidata entry to anchor the brand entity.

Findings by pillar

In a sample report, every finding is shown. In a real free audit, the top three per pillar are visible inline and the rest unlock with your email.

AI Visibility3 findings

Criticalhigh confidenceNeeds action

Homepage exposes 9% of its content without JavaScript

The raw HTML returns 380 characters of body text; the rendered DOM returns 4,210. GPTBot, ClaudeBot, and PerplexityBot do not run JavaScript, so they see roughly 9% of your homepage. Your entire value proposition is invisible to the AI engines.

Recommendation

Server-render the homepage, features, and pricing content. In the Next.js App Router this means Server Components for content blocks rather than "use client". Re-run the audit and confirm the parity score rises above 85.

Methodology: render-parity

Highhigh confidenceNeeds action

SoftwareApplication and Organization schema missing

There is no JSON-LD on the homepage. AI engines cannot confidently identify FlowForge as a specific product entity, so they default to recommending competitors with full SoftwareApplication and Organization markup.

Recommendation

Add Organization schema with sameAs links (LinkedIn, Crunchbase, your Wikidata entry, X) and SoftwareApplication schema with applicationCategory, operatingSystem, and an AggregateRating you can substantiate.

Methodology: schema-softwareapplication

Mediummedium confidenceNeeds action

Docs answers are not structured as extractable passages

Your "How does FlowForge handle SSO" answer is spread across four paragraphs under a generic heading. Answer engines prefer a question-shaped heading followed by a tight 40-to-170-word answer they can lift verbatim.

Recommendation

Restructure your top 20 docs and FAQ entries so each question is an H2 or H3 immediately followed by a single, self-contained answer paragraph. Lead with the answer, then add detail.

Methodology: passage-extractability

Technical2 findings

Highhigh confidenceNeeds action

Interaction to Next Paint is 340ms on the pricing page

CrUX field data shows INP at 340ms (target: under 200ms). A heavy client-side pricing calculator blocks the main thread on every interaction, so the page feels janky when users toggle billing options.

Recommendation

Break up the long JavaScript tasks behind the calculator, defer non-critical work, and yield to the main thread between updates. Consider computing the pricing server-side and hydrating only the interactive toggle.

Methodology: core-web-vitals

Infohigh confidencePassed

HTTPS enforced, HSTS preloaded, no mixed content

TLS is valid, HSTS max-age is set with preload, and no page loads insecure assets. This is the correct baseline.

Recommendation

No action needed.

Methodology: https-and-hsts

Foundational2 findings

Mediummedium confidenceNeeds action

Thin coverage of the "workflow automation" topic cluster

You have 4 articles in the cluster your buyers search; the leading competitor has 31. Cluster-depth analysis puts you in the 2nd quartile, which caps how often you rank and get cited for the category.

Recommendation

Publish 12 articles over the next quarter covering the cluster comprehensively: integrations, use cases by role, comparison content, and BOFU "best workflow automation for X" pages. Interlink them around a pillar page.

Methodology: topical-authority

Infohigh confidencePassed

Titles are unique and within length on every page

All audited pages have a unique title within the 50-to-60 character display range. No duplicates.

Recommendation

No action needed.

Methodology: title-quality

LocalNot applicable

Not applicable: we found no physical location signals on the pages we read, so the local checks do not apply to this site. It did not score.