We Scored 19 Top SaaS Brands for AI Search Readiness. None Passed.

Everyone agrees AI search is changing how people find things. Almost no one has measured whether the sites we all use are actually ready for it. So we did.

We ran 19 of the best-known SaaS and web brands through Citerank's AI Visibility Score and recorded how each one performed. These are companies with large teams, real budgets, and mature marketing. If anyone is ready for AI search, it should be them. The results say otherwise, and the gap they reveal is the opportunity.

The headline: nobody passed

Across 19 brands, the average AI readiness score was 44 out of 100. The median was 45. The single highest score was Zapier at 69, a C. Not one site earned an A or a B. Five reached a C, seven landed a D, and seven scored an F.

Read that again. The most recognizable names in software, and the best of them is a C.

This is the same pattern we see on almost every site we test, but it is more striking when the names are this familiar. Being famous does not make a site machine-readable. The engines do not grade on reputation. They grade on whether they can discover, parse, trust, and act on what is in front of them.

The full data

Here is every site, its overall score and grade, and its five layer scores. Layers are scored 0 to 100.

The agentic web is almost entirely unclaimed

The weakest layer by a wide margin was agentic readiness, which measures whether a site exposes anything an AI agent can call rather than pages built only for human clicking. The average was 18 out of 100. Eleven of the 19 sites scored a flat zero, and only two, Zapier and Semrush, scored above 50.

This is the most important number in the study. AI agents are moving from answering questions to taking actions, and a zero here means an agent can read your marketing page but cannot do anything with your product. For most of these brands there is nothing to call at all. Whoever builds for this first in their category will own it, because right now the field is empty. We wrote the full playbook for this in our guide to agent-ready websites and MCP.

The content is usually fine. The plumbing is not.

Content citability was the strongest shared layer, averaging 60. That matters, because it means the writing itself is often good enough to be quoted. The problem sits upstream of the content.

Discoverability averaged just 38. Schema and entity clarity swung wildly, from a clean 90 down to a zero, with several major brands including Notion, Airtable, GitHub, and Slack scoring zero on schema. When schema is missing, the engine has to guess what your page is and who you are, and guessing is where citations get lost. This is the same discoverability gap we described in the complete GEO guide: the answer is on the page, but the engine cannot cleanly get to it.

The takeaway is encouraging if you are behind. You may not have a content problem. You may have a parsing and discovery problem, and those are faster to fix.

The site that blocked us entirely

One brand, Canva, hard-blocked our crawler and returned nothing, so we left it out of the averages. That is worth calling out on its own. If a site blocks a straightforward crawler, it is very likely blocking AI crawlers too, which means it cannot be read or cited no matter how good the content is. Blocking bots you do not want is reasonable. Accidentally blocking the engines that decide whether you appear in AI answers is not. It is worth checking exactly what your robots rules allow, which we cover in the context of why sites do not get cited.

What this means for you

If brands this large are averaging a 44, the bar in your market is almost certainly just as low. That is good news. You are not trying to out-resource Stripe or Salesforce. You are trying to be more machine-readable than competitors who have done nothing, and most of them have done nothing.

A focused plan gets you ahead quickly:

1. Fix discoverability first. Make sure your real content is in the raw HTML and nothing important is blocked. This was the most common ceiling in the data.
2. Add clean schema. Several major brands scored zero here. Correct Organization, Article, and FAQ markup is one of the fastest wins available.
3. Write for extraction. Your content is probably already decent. Put direct two to three sentence answers under question-style headings so engines can lift them.
4. Claim the agentic layer. Almost no one has. Even a small step here puts you ahead of household names.

How to see where you stand

The 19 scores here came from the same free tool you can run on your own site. Enter your domain in the AI Visibility Score and you will see your overall grade plus which of the five layers is holding you back, in a few seconds and with no login. Most sites, including the famous ones, discover they have far more room than they expected. The ones that act on it now are competing in a nearly empty field.