Key Takeaways
- Read the preference as evidence, not favoritism. Big brands win where the question is vague, because volume of coverage is the only signal available. Nothing in the system is rewarding size for its own sake.
- Watch the advantage collapse as questions narrow. Add a client type, a city, or a specific problem and the national names start dropping out, because they were never described in those terms by anybody.
- Stop competing for the broad question. Best business coach is a question you will not win and your buyer rarely asks in that form. The narrow versions are both winnable and closer to a purchase.
- Expect directories in the middle. When an assistant cannot name a business confidently it names a list instead. A directory in your results is a signal that the category is thinly documented.
- Own one narrow question completely. Being the clearest documented answer to a question fifty people ask beats being the fortieth plausible answer to one that fifty thousand ask.
Ask an assistant a broad question about your category and you will usually get back the same handful of large, well-known names. Ask enough of these and it starts to feel like the system has been built to hand the market to whoever is already winning.
That reading is half right, and the half that is wrong is the useful half. Size correlates with being named, but size is not what is being rewarded, and the difference tells you exactly where a small practice has a chance.
These are field notes: what we see in the shape of real answers in 2026, why the pattern looks the way it does, and which questions are worth your attention as a result.
Do AI assistants prefer big brands?
They prefer whatever is best documented, and big brands are almost always best documented. The preference is real in its effects and misdescribed in its cause. Nothing in the system knows or cares how large you are; it knows how much verifiable material exists about you, and for a national brand the answer is a great deal.
This matters because the two explanations point at completely different responses. If assistants preferred size, a small practice would have nothing to do but grow. If they prefer documentation, a small practice can be better documented than a national brand inside one narrow slice, and that is a project you can actually finish.
The research on influencing generated answers points the same way. The Princeton team behind the generative engine optimization study found that visibility moved with the properties of the source text, such as whether it carried citations, statistics, and quotable authoritative phrasing, rather than with the size of the site it sat on.
Why scale wins the broad questions
Because a vague question has nothing in it to discriminate with. Asked who the best business coach is, an assistant has no client type, no place, no problem, and no budget to narrow on. What is left is the crude signal of how often and how consistently a name appears across everything it has read.
Big brands win that comparison by construction. They have press coverage, review volume, partner pages, conference listings, and years of people writing about them. A four-person consultancy has none of that, and no amount of quality changes the count.
There is also a caution effect. An assistant naming a business is making an implicit claim, and the safest claims are the ones it has seen supported many times. Naming a well-known company is low risk. Naming a practice mentioned twice on the entire public web is not, so it hedges toward the familiar. The same dynamic viewed from inside one niche is in why big fitness apps beat local trainers in AI answers.
No counts and no rates here, because we have none we could publish honestly yet. This is the composition of the answer as the question narrows, and the five rows are illustrative, written to show the pattern rather than to report a measured result.
Written out, because nothing here may live only inside a figure. Asked for the best business coach, the three slots tend to go to two national names and a directory. Add "for small firms" and it is a national name, a directory, and a regional firm. Ask for a coach for agency owners and it becomes a regional firm, a national name, and one individual practice. Add Denver and it is a regional firm and two individual practices. Ask who helps agency owners fix pricing and it is two individual practices and a regional firm. All five rows are illustrative.
Where the size advantage stops mattering
At the point where the question contains something a national brand was never described as doing. That is usually a client type, a place, a specific problem, or a constraint, and it is the single most reliable pattern in any of this.
The reason is mechanical. A national brand is written about in general terms, because that is what generalist coverage produces. It is described as a large coaching company, not as the people who fix pricing for twelve-person creative agencies. When the question uses the second vocabulary, the brand's enormous documentation says nothing relevant and contributes nothing.
| Question adds | What the assistant can now use | Effect on national names | Effect on a solo practice |
|---|---|---|---|
| Nothing, the bare category | Overall coverage volume only | Dominant | Absent |
| A client type | Who each provider is described as serving | Weaker | Possible, if stated on your pages |
| A city or region | Place claims that can be cross-checked | Much weaker | Strong, if anchored outside your site |
| A specific problem | Problem language in the source text | Often absent | Strong, and cheap to earn |
| A constraint, such as budget or format | Terms most coverage never mentions | Absent | Strongest, and least contested |
Read the bottom row twice. The questions where you have the most advantage are the ones almost nobody writes for, because they look too small to bother with. They are also the questions people ask when they are ready to pay somebody.
What a directory in your results is telling you
That the assistant could not confidently name a business. A directory is a responsive answer with no exposure attached: it satisfies the question without asserting that any particular provider is good. When one shows up in your category, treat it as a reading of the category, not of you.
That reading is good news more often than it looks. A directory in the results means the individual providers in your niche are thinly documented, which is exactly the condition under which one well-described practice can take a slot. Crowded categories give you no such opening.
Where it becomes a problem is when the same directory answers every version of the question, including the narrow ones. That usually means nobody in your niche has written the specific page a buyer's question deserves, and whoever writes it first tends to hold it. We covered how that shortlist forms and hardens in why AI answers keep naming the same few coaches.
What a small practice should actually do
Pick one narrow question and become the clearest documented answer to it. Not the best answer, the clearest, because clarity is what a machine can detect and quality is not. That means the client type, the problem, and the outcome written in the buyer's words, on a page about that one thing.
Then get one source you do not own to describe you the same way. This is the step that converts a claim into something an assistant will repeat, and it is the step almost everybody skips in favor of rewriting their homepage for the fourth time.
Do not benchmark yourself against the national names. They are winning a question you are not competing for, and their advantage on it is not reachable by any amount of good work on your site. Google's own guidance on AI features in Search describes answers assembled from pages found across the web, which is a reminder that you are competing page by page, not brand against brand.
Finally, check across more than one engine before drawing conclusions. Perplexity's help center describes an approach built around retrieving and citing current sources, and engines that cite differently produce different name lists. A pattern that holds on one engine and not the others is a quirk. A pattern that holds on all of them is worth acting on, which is the discipline behind what 25-question test runs keep showing us.
Our free check runs 25 real buyer questions from your niche at $0 and emails a plain-English readout within the hour, including which of your questions are held by national names and which are open. It runs as soon as you ask for it.
Questions we hear the most
Do AI assistants prefer big brands?
They favor whatever is best documented, and big brands are usually best documented. It looks like a preference for size because size and coverage travel together, but the thing being rewarded is how much verifiable material exists about you.
Can a solo practice ever beat a national brand in AI answers?
Yes, on narrow questions. Once a question carries a client type, a city, or a specific problem, the national brand often has nothing describing it in those terms, and a well documented solo practice does.
Why does AI name a directory instead of a business?
Because it could not name a business confidently. A directory is a safe answer that is technically responsive, and its appearance usually means the category has too little written about individual providers to choose between them.
Does advertising spend affect who gets recommended?
Not directly. Assistants are drawing on published text rather than ad auctions. Spending does help indirectly over time, by producing coverage, mentions, and reviews that later become the material an assistant reads.
Should I try to rank for broad questions anyway?
Usually not. Broad questions are expensive to compete for, and the buyer asking one is rarely close to hiring. The specific questions are cheaper to win and produce conversations with people who already know what they need.
Do these patterns change between ChatGPT and Perplexity?
The details do. Engines differ in what they retrieve and how much they cite, so the same question can produce different names. The pattern that specificity favors small practices shows up across all of them.
How do I find out which questions I can actually win?
Run the narrow versions of the questions your buyers ask and see who is named. Where you find a solo practice or a regional firm holding slots, the category is winnable. Where every slot is a national brand, move on.
What does the AnswerHalo free check include?
It costs $0. We ask ChatGPT 25 real buyer questions from your niche, then email a plain-English readout within the hour: how many answers named you, who got named instead, and the first fix. It runs as soon as you ask for it.
Your buyers are already asking. Find out what AI tells them.
The free check asks ChatGPT 25 real buyer questions about your niche and sends you the report within the hour, at $0. If it shows you are already getting named everywhere, we will say so in plain words.
Get my free check →