There is a version of this question that gets asked quietly, in the spaces before someone decides to hire an SEO consultant: if this person is so good at this, why doesn't their own website rank?
It's a fair question. And the answer, when you dig into it, is more revealing than most people expect. Not because it exposes fraud, but because it exposes a structural problem with how SEO advice has been packaged and sold for the last decade.
The cobbler's children problem in SEO
The cobbler's children problem in SEO isn't primarily a time problem—it's a credibility problem. It reveals something about what kind of knowledge the practitioner actually has. The cobbler's children have no shoes: it's an old phrase for a pattern that recurs across every service industry, but in SEO it runs deeper than busyness.
A structural engineer who can't make their own building stand up is a different kind of problem than a graphic designer with an outdated portfolio. The building either stands or it doesn't. SEO is closer to the building than the portfolio—and a practitioner who understands the foundations should, in theory, be able to apply them to any site, including their own.
When they can't, or haven't, the question worth asking is: what exactly do they know?
What most SEO advice is actually built on
Most SEO advice is built on a stack that teaches practitioners to operate tools rather than understand infrastructure. The dominant ecosystem centres on a popular CMS, a plugin handling the technical layer, keyword research tools generating the content calendar, and a backlink acquisition strategy built on outreach and guest posting.
This stack has produced millions of pieces of advice, thousands of courses, and an entire consulting industry. It has also produced a generation of practitioners who know how to use the tools without necessarily understanding what the tools are doing—or why.
The plugin manages meta tags, generates sitemaps, flags keyword density, and reports a readability score. All of this is useful. None of it requires you to understand what a meta tag actually communicates, why a sitemap matters, or what keyword density has to do with how a crawler parses a document.
When the tools abstract away the understanding, the practitioner learns to operate the interface. That's not the same as understanding the infrastructure. And when you don't understand the infrastructure, you can't diagnose it when it fails—which is exactly when clients need you most.
The plugin dependency loop
There is a specific dynamic worth naming, because it affects a significant proportion of the SEO consulting market.
Modern SEO plugins have moved considerably since pre-AI times out of necessity. The leading options now generate entity graphs, consolidate schema into a single machine-readable endpoint, offer AI crawler controls, and surface GSC data in the dashboard. These are real capabilities.
The problem isn't what the plugin does. It's what the plugin can't tell you.
A plugin can generate a Person entity; it can't tell you whether the @id reference in that entity is correctly wired to the Article blocks on your content pages, or why the site still isn't cited when it isn't. It can offer a toggle to block AI crawlers; it can't tell you whether the citation crawlers you actually want (OAI-SearchBot, Claude-SearchBot, PerplexityBot) are permitted while training crawlers are blocked, or whether a network-level rule upstream has made the entire robots.txt irrelevant. It can score your content's readability; it can't tell you whether that content is structured for extraction by a language model rather than simply for human reading.
The tool handles the output. What the output means, whether it's correct, and what's missing when the site still doesn't perform, that requires infrastructure knowledge the tool was never designed to replace. And that knowledge is exactly what you're evaluating when you look at a practitioner's own site.
What the absence of an SEO consultant's own results tells you
When an SEO consultant's site doesn't rank for anything meaningful—no non-branded queries, no cluster of articles building topical authority, no evidence of topical traffic beyond their own name—it usually means one of three things.
The first is busyness: the practitioner is so focused on client work that their own site gets nothing. This is the most charitable explanation, and sometimes true. It's also a weak excuse for someone whose value proposition is expertise in this exact area.
The second is that the practitioner is primarily a strategist who relies on others to implement—writers, developers, link builders—and their own site has no team behind it. This explains the gap but raises a different question: if you outsource the implementation, how deeply do you understand what's being implemented?
The third—the one that's most common and least discussed—is that the practitioner knows the surface layer well and the infrastructure layer not at all. They can conduct a keyword audit. They can review a content calendar. They can check a meta description. What they can't do is open the schema, trace the entity graph, check the robots.txt against the AI crawler list, audit whether JavaScript rendering is hiding content from crawlers, explain why a site has 2,000 impressions and zero clicks, or architect a solution for any or all of these problems.
A site that doesn't rank is data. It is the practitioner's own infrastructure, in public view, doing exactly what their advice should prevent.
What to actually look for when hiring an SEO consultant
The question "does this person's own site rank?" is a useful starting point but imprecise. A newer site won't rank yet regardless of how well it's built. A practitioner in a brutally competitive niche may be outranked while still being excellent. What you're really asking is: does this person's site demonstrate the infrastructure understanding they're claiming?
That's a more specific question, and it has specific answers.
Look at the schema. Does the site have a correctly implemented entity graph—Person linked to ProfessionalService, Article blocks with author references, FAQPage schema on content pages? Or does it have the default schema that shipped with the plugin and was never customised?
Look at the content structure. Does each article open with a direct, self-contained answer in the first paragraph? Is the content structured to be extracted by an AI engine, or written to guide a human reader through a narrative?
Look at the robots.txt. Does it explicitly name and allow the major AI crawlers—GPTBot, OAI-SearchBot, ClaudeBot, Claude-SearchBot, PerplexityBot? Or does it say User-agent: * and leave everything to inference?
Check their organic footprint publicly. You don't need anyone's permission to do this. Enter the consultant's domain into any SEO tool with a domain overview feature; Semrush, Ubersuggest, and Ahrefs all offer free or limited access, and you'll see their estimated organic traffic, the keywords they rank for, and crucially, the split between branded and non-branded traffic. Branded traffic means people searching their name. Non-branded traffic means people finding them through the topics they claim to know. A practitioner whose organic visibility is almost entirely branded (their name, their business name, and nothing else) is not demonstrating the expertise they're selling. This is public information. It takes two minutes to check.
These are not trick questions. They are the basic infrastructure checks that any practitioner who understands the current landscape should be able to answer for their own site immediately.
A fair objection here: these are different disciplines. SEO, AEO, content strategy, entity infrastructure, schema implementation, AI crawler configuration—in most agencies and consultancies, they're handled by different people. A strategist, technical SEO expert, content writer, and a developer. The brief gets translated between them. Things get lost. The infrastructure layer, which nobody owns clearly, tends to fall through the gaps.
The rare thing isn't expertise in one of these areas. It's someone who holds all of them and can move between them without translation loss; who designs the entity graph and writes the article that populates it, who configures the robots.txt and understands why the content structure above it matters, who can audit what's broken and build what replaces it. That's not a team. It's a different model entirely. And it's the model that closes the gap between surface advice and infrastructure results.
Infrastructure doesn't lie
The most useful due diligence you can do before hiring anyone for this work is to look at what they've built for themselves.
Not their client list, their certifications, or their follower count. Those are claims. A site is evidence.
A practitioner who understands infrastructure has built it. Their entity graph is wired. Their AI crawler access is explicit. Their content is structured for extraction, not just for reading. Their schema validates at zero errors. Their non-branded organic footprint, visible to anyone in thirty seconds via a domain overview tool, shows a site earning traffic on topics, not just on its own name.
That site exists. It's public. You can look at it.
If the practitioner you're evaluating can show you that their own site demonstrates the principles they're recommending, you're talking to someone who knows the difference between operating a tool and understanding the infrastructure the tool is supposed to serve.
If they can't, and the site hasn't been touched in months, runs on default plugin schema, and ranks for nothing but their own name, the question to ask isn't "why haven't they had time?" It's "what, exactly, are they going to build for me?"
Why don't SEO consultants rank their own websites?
SEO consultants often don't rank their own sites for one of three reasons: they're too focused on client work to invest in their own infrastructure, they rely on outsourced implementation and lack deep hands-on knowledge, or they understand the surface layer of SEO (keywords, meta tags, plugin settings) without understanding the underlying infrastructure that determines real discoverability. The third reason is the most common and the most relevant when evaluating who to hire.
How can I tell if an SEO consultant actually knows what they're doing?
Look at their own site, not their testimonials. Does it have a correctly wired schema entity graph? Does it rank for non-branded queries? Does the robots.txt explicitly allow AI crawlers? Is the content structured for extraction, not just for human reading? Run their domain through a free SEO tool like Semrush, Ubersuggest, or Ahrefs. Does the branded/non-branded split show they're earning traffic on topics, not just their name? A practitioner who understands infrastructure has built it—on their own site, in public view. That's more reliable evidence than any case study or certification.
What is the difference between surface SEO and infrastructure SEO?
Surface SEO covers the things a plugin or tool can check and fix: meta tags, keyword density, readability scores, sitemap generation. Infrastructure SEO covers the things that determine whether a site is actually discoverable to search engines and AI systems: entity graph implementation, schema markup wired with @id references, AI crawler access via robots.txt, JavaScript rendering that doesn't hide content from crawlers, and content structured for AI extraction. Most SEO advice addresses the surface. Most visibility problems live in the infrastructure.
Does a plugin handle SEO properly?
An SEO plugin handles the surface layer well via meta tags, sitemaps, and basic schema. What it doesn't handle is the infrastructure layer that increasingly determines AI search visibility: entity graph construction, AI crawler access controls, schema wired with @id references across entity types, and content structured for extraction by language models. A plugin can tell you your meta description is too long. It cannot tell you whether your site is visible to ChatGPT, Claude, or Perplexity.
What should I look for before hiring an SEO consultant?
Before anything else, run their domain through a free SEO tool—Semrush, Ubersuggest, or Ahrefs all offer basic domain overviews at no cost. Look at the branded vs non-branded traffic split: a practitioner whose organic visibility is almost entirely their own name is not demonstrating the expertise they're selling. Then check the infrastructure: does the site have a correctly implemented schema entity graph with Person and ProfessionalService entities linked via @id? Are AI crawlers explicitly permitted in robots.txt? Is the content structured answer-first? A practitioner who advises on discoverability but hasn't built it for themselves is worth questioning carefully.
Aimee Q Devlin is a Systems and Infrastructure Architect based in San Miguel de Allende, Mexico. She works with founders and operators of established businesses who are ready to rebuild their systems properly—including the infrastructure that makes those systems discoverable. The Infrastructure Audit is where most engagements begin.