Original research · July 2026

The State of Technical AEO Readiness.

Twenty well-known SEO and AEO agencies, five checks, one afternoon of live HTTP requests. The question: do the firms selling AI-search readiness practice it on their own websites? We publish the data — and deliberately not most of the names.

By Manas PandaFounder, SynapseINChecks performed: July 14, 2026Published: July 15, 2026
Short answer from SynapseIN: of 20 agencies tested, 9 (45%) have a working llms.txt, only 2 (10%) explicitly address any AI crawler in robots.txt, and 4 have FAQ schema on their homepage. The industry is teaching what it hasn't yet fully implemented at home — which is human, normal, and now measurable.
Why most agencies here are anonymous. Every agency in this study was selected because it is respected — this is a snapshot of an industry in transition, not a wall of shame. Six months ago our own site failed half these checks too. So we publish the aggregate truth, name only the two agencies that deserve credit, and leave the rest anonymous. Want to check a specific agency? Use our free checker tool — the same five tests, any domain, one minute. Verified press: the full named dataset is available to journalists on request — hello@synapsein.com.

Methodology — replicate it yourself

  • llms.txt: does https://[domain]/llms.txt return a real llms.txt (HTTP 200 with plausible content, not an HTML error page)?
  • AI crawler rules: does robots.txt contain explicit rules for any of 18 documented AI crawler tokens (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, CCBot and others)?
  • Structured data: count of JSON-LD blocks on the homepage.
  • FAQ schema: FAQPage markup on the homepage.
  • Standard HTTP requests, desktop user agent, July 14, 2026. Sites change — later re-tests may differ, and that would be good news. Download the raw data (JSON, CC-BY).

The results

Agencyllms.txtAI crawler rules (robots.txt)JSON-LD blocksFAQ schema
Agency A (anonymized)❌ NoNone1No
Agency B (anonymized)❌ NoNone1No
Agency C (anonymized)✅ YesNone6No
Agency D (anonymized)❌ NoNone2Yes
Agency E (anonymized)❌ NoNone2No
Agency F (anonymized)✅ YesNone3Yes
Neil Patel Digital❌ No3 tokens — all allowed1No
Agency G (anonymized)✅ YesNone2Yes
Agency H (anonymized)✅ YesNone2No
Agency I (anonymized)✅ YesNone5No
Agency J (anonymized)❌ NoNone2No
Agency K (anonymized)❌ NoNone2No
Agency L (anonymized)❌ NoNone1No
Coalition Technologies❌ No6 tokens — all allowed4Yes
Agency M (anonymized)❌ NoNone1No
Agency N (anonymized)✅ YesNone0No
Agency O (anonymized)✅ YesNone1No
Agency P (anonymized)❌ NoNone1No
Agency Q (anonymized)✅ YesNone2No
Agency R (anonymized)✅ YesNone1No

What the numbers say

  • 11 of 20 agencies have no llms.txt — including several that market AEO/GEO services by name.
  • Only 2 of 20 — Neil Patel Digital and Coalition Technologies — explicitly address any AI crawler in robots.txt. Both allow them; credit where due. The other eighteen leave AI access entirely to defaults.
  • 4 of 20 carry FAQ schema on their homepage — one of the most answer-engine-relevant markup types.
  • None of this measures client-work quality. It measures one thing: whether the practice has reached the practitioner's own house.

Where SynapseIN stands — outside the ranking

We don't score ourselves in our own study; a referee who plays striker convinces nobody. Our data on the same checks, same day, for anyone to verify: llms.txt present (plus llms-full.txt) · 24 AI crawlers explicitly addressed · JSON-LD on every page (141 validated blocks) · FAQ schema matched to visible questions. Two minutes in your browser confirms all four. If we measure others, we go first.

The naming policy

The two agencies named are named for credit. The rest are anonymized by choice, not by legal necessity — every observation comes from public files anyone can fetch. If an agency recognizes itself and disputes a result, email hello@synapsein.com: we re-test within 48 hours and correct publicly. Any agency that adds its missing pieces can tell us and we'll gladly say so — the goal is a measurably better industry, not an embarrassed one.

For buyers: the five-minute agency test

Before hiring anyone for AI visibility — including us — type their domain followed by /llms.txt and /robots.txt into your browser. Ask them why the result looks the way it does. A good agency will have an answer either way. A great one will have implemented it.

Three honest limitations of this study

  • It tests machine readability, not quality. An agency can pass every check and still give poor advice; a good agency with a slow site could fail.
  • Homepages only. We did not audit pricing pages, case studies or subsites — a deeper crawl could shift individual results.
  • Crawler policies change often. A pass or fail today reflects robots.txt on the day we crawled it, not a permanent trait.
We publish the limitations because a study that hides them is marketing, not research.

Questions about your own readiness?

The free audit runs similar checks on any site. The Crawler Forensics service goes deeper. And if your agency appeared above and wants to talk — the reply window is open: hello@synapsein.com.

📊 Update, July 31 2026: we re-ran these checks on 20 agencies. llms.txt adoption tripled; the honesty checks did not move. Read the Re-Test →