By Lesli Rose · May 13, 2026 · 7 min read
This agency had built something real. Enterprise partner of the year from one of the web's largest development platforms. A client roster including a major social network, a wellness app, and a fintech company. Named as an early designee in an enterprise AI development program. Four years old, growing at over 70% annually. Their commercial flagship was AEO -- Answer Engine Optimization, the practice of making content citeable by AI systems like ChatGPT, Perplexity, and Google's AI Overviews.
This is an agency that sells AI citability as a service. When I ran the audit, their robots.txt was among the best AI bot configurations I have seen on any agency site.
User-agent: GPTBot Allow: / User-agent: ChatGPT-User Allow: / User-agent: OAI-SearchBot Allow: / User-agent: anthropic-ai Allow: / User-agent: ClaudeBot Allow: /
Ten AI crawlers explicitly named. Full access granted. No accidental blocking. It is the kind of robots.txt configuration I recommend to every client I work with. Then I checked their llms.txt.
HTTP/2 404
The infrastructure door is wide open. There is nothing on the other side for AI systems to read.
Any business missing an llms.txt is leaving AI citation potential on the table. For a general-purpose Webflow agency, it is a gap. For an agency whose sales page leads with "we make your content citeable by AI" -- it is a credibility problem that will surface in every sales conversation with a sophisticated buyer.
The buyers evaluating an AEO agency are exactly the buyers who will run these checks. They will curl the llms.txt. They will check whether the agency's own /aeo-agency page appears in AI responses to "best AEO agency." They will notice that the agency's Organization schema is missing the sameAs links that connect the website entity to its LinkedIn and YouTube profiles -- the same sameAs links the agency would add to a client's site on day one of an engagement.
An llms.txt for this agency would take two hours to write and deploy. It is the fastest credibility move available. It is also the clearest proof of approach.
The homepage has two schema types: a partial Organization block and a WebSite block. The Organization schema is missing three things that matter most for AI entity recognition: sameAs links to LinkedIn and YouTube, the founding year, and a company description field.
Without sameAs links, AI knowledge graphs treat the website entity and the LinkedIn company page as potentially different organizations. The agency's LinkedIn following, post engagement, and professional activity contribute zero to the domain entity's citation authority until the two are connected.
Missing from the site entirely: Service schema on the AEO and SEO service pages, Person schema for the founding team, FAQPage schema on any page, BreadcrumbList sitewide, and AggregateRating (which requires a base schema layer before it can be added). Six schema types absent from a site that is actively selling schema implementation as a client deliverable.
The /aeo-agency page is the commercial flagship. It targets "AEO agency" and "answer engine optimization agency" -- queries that are low-competition today but growing fast as brands realize traditional SEO does not automatically translate to AI citation. The page has good content. There is a six-step methodology laid out clearly, well-written explanations of what AEO produces and why it matters, and pricing signals ($4k+ one-time, $5-15k/month).
The machine-readability of that page is near zero. No Service schema. No FAQPage schema. No structured JSON-LD of any kind on the page. AI systems parse the content as marketing copy, not structured knowledge. The six-step methodology -- the exact kind of factual, numbered, extractable content that AI systems love to cite -- is invisible to structured data extraction.
Adding FAQPage schema around that methodology is a two-hour task. No new content needed. The answers already exist. This is the highest-ROI single action available on the site.
The homepage meta description read as follows during the audit:
"We launch sites, fast. [Agency name] is an award-winning Webflow agency offering custom development, site migration, ongoing maintenance and SEO services."
That description could belong to any of the several hundred Webflow agencies currently active. It contains nothing that flags the AEO service, the enterprise partnership award, the named client roster, or the AI development program designation. Every time the homepage URL appears in a search result, the SERP listing reinforces "another Webflow agency" instead of "the Webflow agency with an AEO practice."
This is a fixable problem in under ten minutes. The differentiation exists. It is just not in the meta description.
The content library has solid awareness coverage -- explanations of what AEO is, why it matters, what AI search is doing to organic traffic. What is missing is the content that converts an aware buyer into a booked call.
Queries like "how to choose an AEO agency," "what does an AEO retainer include," and "[agency name] vs [competitor]" had no matching pages. These are the searches that happen during a multi-vendor evaluation -- the phase where buyers narrow from five options to two. The agency was showing up well in awareness queries and disappearing during the decision phase.
Three decision-stage pages targeting AEO evaluation queries -- written with Service schema and linked from /aeo-agency -- would close that gap and give the page cluster enough internal authority to compete on the commercial terms.
62
Technical SEO
71
On-Page SEO
18
Schema
44
AI Discoverability
The earned visibility score (64) reflects real authority. Multiple industry award mentions indexed on third-party domains. Named client case studies with recognizable logos. Active on four social platforms. The brand has done the work to build external credibility and it shows in the signals AI systems use to evaluate citation worthiness.
On-page SEO at 71 reflects disciplined title tag and heading work across the main service pages. The AEO and SEO pages have clear H1s, keyword-relevant titles, and descriptive URLs. Technical SEO at 62 reflects a solid Webflow foundation -- HTTPS valid, sitemap declared in robots.txt, Cloudflare CDN active, mobile rendering enabled. The Webflow HTML payload is verbose (706KB), but Cloudflare compression mitigates the worst of it.
The AI crawler policy in robots.txt is legitimately the best I have seen. Every major AI crawler is explicitly named and granted access. The infrastructure discipline is there. The content waiting for those crawlers is not yet optimized to take advantage of it.
This agency's competitive position is unusual. No other agency in the market holds the combination of Webflow build authority and AEO as a named, priced service. That is a defensible moat. The risk is that larger B2B content agencies add Webflow capability, or that Webflow-native agencies start offering AEO, before this agency cements the position in AI search.
The window to become the entity that AI systems cite when someone asks "which agency should I hire for AEO?" is open right now. The three changes that would move the needle most -- llms.txt deployment, FAQ schema on /aeo-agency, and completing the Organization schema sameAs links -- are a combined eight hours of work.
The irony is that this agency knows exactly what those three things are. They are the same things they would deliver to a client on day one.
Is Your Agency in the Same Position?
Agency sites are often the last to get the work that agency clients get first. I run a free visibility report that shows exactly what AI systems see when they crawl yours.
Run Your Visibility ReportAn llms.txt file sits at the root of your domain and tells AI crawlers which pages represent your organization, what you do, and which content should be weighted for citation. It is the document AI systems read after robots.txt lets them in. Without it, AI crawlers decide on their own framing -- usually defaulting to third-party sources over your own site.
robots.txt controls access. llms.txt controls framing. A best-practice robots.txt opens the door for AI crawlers. An llms.txt gives them a map of what they just walked into. You can have one without the other, but both are needed for full AI discoverability.
AI systems build entity pictures by correlating signals across sources. sameAs links connect your website entity to your LinkedIn, YouTube, and other verified profiles. Without them, your social presence and your website contribute authority independently rather than compounding. A business with 10,000 LinkedIn followers and no sameAs link gets zero LinkedIn authority credited to the domain entity.
Not perfect -- but the gaps that exist on an AEO agency's own site are exactly the gaps sophisticated buyers will check. The fastest credibility move for any agency selling AI citability is to be the proof, not just the pitch.