Original Research
SaaS AEO Readiness Study 2026: We Scanned 50 Leading SaaS Homepages for AI Search
Published July 4, 2026 by vito.wu · Data collected 2026-07-04
How ready are the world's best-known SaaS companies for AI search? We ran the homepages of 50 leading SaaS products through the same open AEO Checker engine that powers this site — scoring each page across 10 dimensions of answer engine readiness, from structured data to content extractability. 44 pages could be analyzed; 6 blocked automated access. The results challenge the assumption that big-brand marketing sites are automatically AI-ready.
Key Findings
- The average score was 64.3/100— solidly "good" but far from AI-ready, with only 10 of 44 sites (23%) reaching an excellent grade.
- Only Basecamp cleanly passed the page-weight check. Of the 44 homepages measured, 27 exceeded the healthy raw-HTML budget outright and 16 landed in the warning zone — bloated markup is the industry norm.
- Only 30% show a visible date and only 30% expose author/source signals — the two credibility signals AI engines lean on most are among the most neglected.
- Several famous products are nearly invisible to AI crawlers. The lowest scorers were Miro (10/100), Mixpanel (20/100), GitHub (35/100) and Notion (35/100) — pages that render largely client-side leave AI crawlers an empty shell.
- Dev-tool and SEO companies lead. Atlassian and Netlify topped the table at 90/100, with Semrush, Ahrefs, Datadog and Auth0 at 85 — teams closest to the crawler ecosystem score best on it.
Methodology
On 2026-07-04 we fetched the public homepage HTML of 50 well-known SaaS products (selected for brand recognition across CRM, productivity, dev tools, analytics, and marketing categories; one homepage per company domain) with a standard desktop browser user-agent, then scored each page with the open-source analysis engine behind our AEO Checker — the exact same code any visitor can run against their own site.
The engine evaluates 10 checks in four groups: content structure (Q&A structure, H1 clarity, paragraph extractability), structured data (schema presence, type coverage, field completeness), technical accessibility (mobile viewport, page weight), and credibility (visible dates, author/source signals). Each check returns pass, warn, or fail; the weighted total produces a 0-100 score and a grade. Pass rates below count only clean passes — warnings are not passes.
Honest caveats: 6 of 50 sites (Zendesk, Canva, Dropbox, Moz, Hootsuite, GitLab) blocked automated fetching and are excluded from all statistics — percentages use only the 44 analyzable sites as the denominator. Per-check pass rates use only sites where that check produced a result. This study measures the server-rendered homepage HTML— the same view most AI crawlers get — not the fully hydrated in-browser experience. A homepage is also just one page; a company's docs or blog may score very differently.
Pass Rates by Check: Where SaaS Homepages Fail
Sorted from worst to best. The bar shows the share of analyzable sites that passed each check.
Page Load Performance(1/44 sites)
2%Is the raw HTML payload within a healthy size budget for fast parsing?
Schema Type Coverage(13/44 sites)
30%Are the right schema types (Organization, FAQPage, Article…) used for the content?
Publication Date Visibility(13/44 sites)
30%Can a crawler find a visible published/updated date on the page?
Author/Source Signals(13/44 sites)
30%Does the page expose author or organizational source signals AI can attribute?
H1 Title Clarity(26/44 sites)
59%Does the page have a single clear H1 stating what it is about?
Q&A Structure Detection(30/44 sites)
68%Does the page contain question-and-answer shaped content AI can quote directly?
Key Field Completeness(30/44 sites)
68%Do the schema objects include the required fields, not just a bare @type?
Schema Markup Presence(32/44 sites)
73%Does the page ship any JSON-LD structured data at all?
Paragraph Extractability(40/44 sites)
91%Are paragraphs short and self-contained enough for AI extraction?
Mobile Viewport Configuration(44/44 sites)
100%Is a responsive viewport declared so crawlers treat the page as mobile-ready?
Full Results: All 44 Analyzed Homepages
| # | Site | Score | Grade |
|---|---|---|---|
| 1 | Atlassian | 90 | excellent |
| 2 | Netlify | 90 | excellent |
| 3 | Semrush | 85 | excellent |
| 4 | Ahrefs | 85 | excellent |
| 5 | Datadog | 85 | excellent |
| 6 | Auth0 | 85 | excellent |
| 7 | Klaviyo | 80 | excellent |
| 8 | Amplitude | 80 | excellent |
| 9 | Okta | 80 | excellent |
| 10 | Zapier | 80 | excellent |
| 11 | HubSpot | 75 | good |
| 12 | Figma | 75 | good |
| 13 | Shopify | 75 | good |
| 14 | Asana | 75 | good |
| 15 | Intercom | 75 | good |
| 16 | Twilio | 75 | good |
| 17 | Salesforce | 70 | good |
| 18 | Stripe | 70 | good |
| 19 | monday.com | 70 | good |
| 20 | ClickUp | 70 | good |
| 21 | Box | 70 | good |
| 22 | DocuSign | 70 | good |
| 23 | Loom | 70 | good |
| 24 | Webflow | 70 | good |
| 25 | Basecamp | 70 | good |
| 26 | Buffer | 65 | good |
| 27 | Typeform | 65 | good |
| 28 | Zoom | 60 | good |
| 29 | Airtable | 60 | good |
| 30 | Mailchimp | 60 | good |
| 31 | Calendly | 60 | good |
| 32 | SurveyMonkey | 60 | good |
| 33 | PagerDuty | 55 | fair |
| 34 | Vercel | 55 | fair |
| 35 | Grammarly | 50 | fair |
| 36 | Slack | 45 | fair |
| 37 | Segment | 45 | fair |
| 38 | Supabase | 45 | fair |
| 39 | 1Password | 45 | fair |
| 40 | Linear | 40 | fair |
| 41 | Notion | 35 | poor |
| 42 | GitHub | 35 | poor |
| 43 | Mixpanel | 20 | poor |
| 44 | Miro | 10 | poor |
Scores reflect the server-rendered homepage HTML on 2026-07-04 as measured by the AEO Tools checker. Sites change constantly — run any URL through the free checker for a current result. Excluded (blocked automated access): Zendesk, Canva, Dropbox, Moz, Hootsuite, GitLab.
What This Means for Your Content
The most striking pattern is that the failures are concentrated in the cheapest fixes. Visible dates (30% pass) and author signals (30% pass) cost minutes to add, yet most of the industry skips them — while AI engines increasingly use exactly these signals to decide which sources to trust and cite.
The second lesson is that client-side rendering is an AEO tax. The lowest scorers in this study are not badly built sites — they are sophisticated JavaScript applications whose server response contains almost nothing for a crawler to read. If an AI engine's crawler sees an empty shell, your brand does not exist in its answer.
If you want to avoid the same gaps, work through the AEO Checklist & Best Practices — items 4-6 cover exactly the date, author, and schema signals this study found missing — or start by scoring your own page with the free AEO Checker.
Frequently Asked Questions
Q: Can I reproduce these results?
Yes. Every score comes from the same engine behind the free AEO Checker on this site — paste any of the URLs above into it and you will get the current per-check breakdown. Sites update their homepages constantly, so exact scores will drift over time.
Q: Does a low score mean the company is bad at SEO?
No. This study measures answer-engine readiness of one page — the homepage — not overall SEO. Several low scorers rank superbly in traditional search. The point is that traditional SEO strength does not automatically translate into AI-extractable content.
Q: Why were some sites excluded?
Their servers blocked our fetch (HTTP 403 or connection failures) — a bot defense, which is their right. We report them as unanalyzable rather than guessing at scores.
Q: Can I cite this data?
Yes — the dataset is published under CC BY 4.0. Cite "AEO Tools, SaaS AEO Readiness Study, July 2026" and link to this page.
How does your site compare?
Run the same 10-dimension analysis on any URL — free, no signup.
Check Your AEO Score →