Ask ChatGPT to recommend a business coach in your city. Then ask Perplexity. Then run the same question through Google and watch the AI Overview load before the blue links do.
Somebody gets named in those answers. Right now it probably isn’t you.
That’s the shift most coaches and consultants have not adjusted to. For fifteen years the game was ranking a page. Now a machine reads the whole web, writes a short answer, and cites three or four sources. If you’re not one of them, you don’t get the click, the call, or the deal. You don’t even get seen.
Here’s what makes this urgent. In August 2026 Google updated its spam policy language to cover “attempting to manipulate generative AI responses in Google Search.” So the shortcuts are already being policed. The window for gaming this closed before most people knew it opened. What’s left is the boring, durable work. That’s actually good news for you, because it’s work most of your competitors won’t do.
Why AI Search Broke Your Old Content Strategy
Traditional SEO rewarded pages. AI search rewards claims.
When a language model builds an answer, it isn’t ranking ten documents and handing you a list. It’s assembling a response from statements it found across many sources, and then citing the sources it trusted most for those specific statements. The unit of value moved from the page to the sentence.
That changes what good content looks like. A 2,000-word article that circles a topic without ever stating a clear, quotable position gives the model nothing to lift. A 900-word piece with six sharp, specific, attributable claims gives it plenty.
It also changes who wins. Big domains still have an advantage, but it’s smaller than it was in classic search. Models care a lot about whether a claim is specific, consistent across sources, and attached to a real person with real credentials. A solo consultant with a clear point of view and a well-structured site can absolutely get cited next to a national publication. I’ve watched it happen.
The catch is that most coaching and consulting content is written to sound impressive rather than to say something. Vague authority does not get cited. Specific expertise does.
The Five Things AI Engines Look For
After watching what actually gets pulled into AI answers in this space, the pattern is consistent. Five things show up over and over.
First, a direct answer near the top. Not a story, not a warm-up, not three paragraphs about how the industry is changing. State the answer in the first 60 words, then support it. Models grab the clean statement.
Second, specificity you can defend. “Most coaches undercharge” gets ignored. “Coaches who move from hourly to a fixed 90-day engagement typically raise their effective rate by 40 to 60 percent because they stop selling time and start selling a finish line” gets cited. Numbers, timeframes, named mechanisms.
Third, structure the machine can parse. Real H2s that pose the question a person would ask. Short paragraphs. Lists where lists make sense. A FAQ block at the bottom with the exact questions people type. This is not decoration. It’s how the content gets chunked and retrieved.
Fourth, a visible author with real credentials. Models weight who said it. An article with no byline, no bio, and no link to a credible profile is a weaker citation than the same article signed by someone with a track record the model can verify elsewhere on the web.
Fifth, corroboration. If your claim shows up only on your site, it’s a lone signal. If a version of it also appears in a podcast transcript, a guest article, a LinkedIn post, and a directory bio, the model sees a consistent entity making a consistent claim. That’s the piece almost nobody works on deliberately.
How to Structure a Page So It Gets Pulled
Take a topic you actually know cold. Say it’s how consultants should price a retainer. Here’s the build.
Open with the answer. One paragraph, under 60 words, that a model could quote standalone. “A consulting retainer should be priced against the client’s outcome, not your hours. The most reliable structure is a fixed monthly fee tied to a defined scope of deliverables, set at 8 to 12 percent of the annual value the engagement is expected to create.”
Then break the body into questions. “What should a consulting retainer include?” “How long should a retainer run?” “When should you raise a retainer fee?” Each one is a query someone actually types. Each H2 becomes a retrievable chunk.
Under each question, answer in the first two sentences. Support it after. Do not save the payoff for the end. Models often only read the top of a section.
Add one concrete example with numbers. Real or realistic, but specific. “A marketing consultant charging $2,000 a month for 20 hours of work is billing $100 an hour. The same consultant delivering a defined pipeline system for $6,500 a month is billing for the outcome, and the work takes about 14 hours.” That paragraph is the kind of thing that gets quoted.
Close with a FAQ of three to five questions in plain language. Answer each in 40 to 80 words. This is the single highest-leverage block on the page for AI citation and featured snippets both.
Add author schema and Person schema. If your site doesn’t currently tell search engines who wrote the piece and why they’re qualified, you’re leaving a required signal on the table.
Build the Entity, Not Just the Page
This is the part that separates people who get cited occasionally from people who get cited consistently.
Language models don’t just index pages. They build a picture of entities. A person, a company, a concept. If “Gilberto Herrera” or your name is a clearly defined entity with a consistent description across dozens of sources, the model has something to attach a recommendation to. If your name means nothing, or means five different things depending on where it appears, you get skipped.
So make your description consistent everywhere. Same one-line positioning on your site, LinkedIn, your podcast bio, your speaker page, your Google Business Profile, every guest post byline. Same core keywords. Same business names.
Then get mentioned in places that aren’t yours. Podcast interviews with published show notes. Guest articles on industry sites. Quotes in journalist queries. Association member directories. Conference speaker listings. These are the corroborating sources that turn a self-claim into an established fact.
And keep publishing on a schedule. Freshness matters more in AI retrieval than it did in classic ranking, because models favor recent sources when a topic moves fast. A site that published four solid pieces last month reads as active. A site with a 2024 blog reads as abandoned.
What Not to Do
A few things will actively hurt you now.
Don’t write content engineered purely to be scraped. Thin pages stuffed with quotable claims and no substance are exactly what the August 2026 spam update language is aimed at. It’ll work briefly and then cost you.
Don’t publish unedited AI drafts. Models are increasingly good at recognizing generic, low-signal text, and so are readers. If your article could have been written about any industry by anyone, it has no citation value.
Don’t chase every keyword. Ten pages of deep, specific, genuinely expert content on the things you actually do will outperform sixty shallow pages on everything adjacent. Depth is a trust signal.
And don’t ignore your own website structure while chasing off-site mentions. If your site is slow, your pages have no schema, and nobody can tell who wrote what, the mentions have nowhere good to point.
The 30-Day Version
If you want a starting point that actually fits into a working week, do this.
Week one, pick your four core topics. The ones you’d want to be known for. Audit what you already have against them.
Week two, rewrite your best existing page using the structure above. Answer up top, question-based H2s, one specific numeric example, a real FAQ block, author bio and schema.
Week three, standardize your entity. Same bio, same positioning line, same links across every profile you own. Fix any inconsistencies. This takes an afternoon and most people never do it.
Week four, get one external mention. One podcast, one guest post, one journalist response. Then repeat that every month.
That’s it. It isn’t clever. It compounds.
The Real Point
AI search didn’t kill content marketing. It killed vague content marketing.
If you have real expertise and you’re willing to state it plainly, with specifics, in a structure a machine can read, you’re in better shape than you’ve been in years. The people who lose are the ones producing volume without a point of view.
Your expertise is the asset. The work is making it legible.
Your Next Step
If your content is getting traffic but not clients, or getting neither, the problem is usually upstream of the writing. It’s positioning, offer, and flow. That’s what we fix inside the Profitable Pro Accelerator. We build the full Profit Flow so attention actually turns into premium clients instead of just impressions. Apply at gilbertoherrera.com.
Frequently Asked Questions
How long does it take to get cited by AI search engines?
Most sites see first citations in 60 to 120 days after they fix structure and start publishing consistently. It moves faster if your name already appears on third-party sites. It moves slower if your site has no author markup or your entity description is inconsistent across profiles.
Does AI search optimization replace SEO?
No. It sits on top of it. The technical basics still matter, including speed, crawlability, internal links, and schema. What changes is the writing. You now optimize for quotable, specific claims inside a clear structure, not just keyword placement.
Can a small coaching business get cited next to big publications?
Yes, more often than in classic search. Models weight specificity and author credibility heavily. A solo consultant with a defined niche, real credentials, and consistent mentions across the web can outrank a generic article from a large site on a narrow question.
What is the single biggest mistake coaches make with AI search?
Writing generically. Content that could apply to any business in any industry gives a model nothing to cite. Specific numbers, named frameworks, and defensible claims are what get pulled into answers.
Do I need to publish more content or better content?
Better, then more. Four deep, specific pieces a month on topics you genuinely own will outperform twenty shallow posts. Consistency matters, but depth is what earns the citation.
Internal Links for This Article
- Link to “Why Your Coaching Website Traffic Dropped in 2026” at gilbertoherrera.com/coaching-website-traffic-drop-2026 using anchor text: “why your coaching website traffic dropped”
- Link to “Google Business Profile for Coaches” at gilbertoherrera.com/google-business-profile-for-coaches using anchor text: “set up your Google Business Profile”
- Link to “How to Write a Client Case Study That Sells” at gilbertoherrera.com/how-to-write-a-client-case-study using anchor text: “client case studies that get cited”
