Key Takeaways
- Expect a longer run to reach further down the list. Google says Deep Research can browse up to hundreds of websites for a single question. A page that never surfaces in a fast answer can still be read on the twelfth pass.
- Write for the sub-question, not the headline question. A long run breaks one question into a plan of smaller ones. Each of those is a separate chance to be the page that answers it.
- Stop assuming the roundup articles win by default. They win the quick answer because they are the fastest way to name several businesses at once. Depth is where individual websites and professional registers start to count.
- Make your own site say the things nobody else will say for you. Price, format, who you work with, where you are. A long run rewards a source that settles a detail, and nothing settles your details but you.
- Treat this as one more reason to be findable, not a new tactic. There is no Deep Research setting to optimize. The work is the same work: be on the web, be specific, and be confirmable somewhere other than your own site.
A prospective client no longer has to settle for the first three names an assistant offers. They can ask it to go away, read for a few minutes, and come back with a written report.
That is what Deep Research is, and the obvious question for anyone running a small practice is whether it changes who ends up named. It does, though not in the direction most people assume.
Here is the short answer, what a long run does differently, who it turns up, how the two kinds of answer compare, and what is worth doing about it this week.
Does Gemini Deep Research change who gets recommended?
It changes who gets read, which changes who can be recommended. A quick answer works from a handful of sources. A long run works from far more, and the extra sources are where the small and specific businesses live.
Google's own description of the feature is worth reading rather than paraphrasing. Its Deep Research overview says the feature can automatically browse up to hundreds of websites on your behalf, reason over what it finds, and produce a multi-page report in minutes. The same page describes the three stages: it turns the prompt into a research plan, searches and browses against that plan, then writes the report.
Nothing in that is a ranking system you can game. It is a longer read. The practical consequence is that being outside the first handful of sources stops being fatal, and being vague stops being survivable.
What a deep run actually does differently
Three things: it plans before it searches, it searches many more times, and it writes at length instead of in a paragraph. Each one changes what kind of page is useful to it.
The plan matters most. A quick answer takes your buyer's question as typed. A long run breaks it into smaller questions first, and Google says you can see that plan and edit it before the run starts. So the question your page has to match is no longer "best nutrition coach near me". It is one of six or eight narrower questions nobody typed out loud.
The length of the report matters too. A three-name answer has room for three names. A multi-page report has room for a section on how people in the field are usually credentialed, a section on typical pricing, and a section on who works with a particular kind of client. Those sections need sources that settle details, which is a different job from a source that lists names. What a page has to do to be quoted at all is covered in what grounding means in AI answers.
Who a longer search turns up
The businesses that show up only in the long version are the ones with a real presence and no listicle coverage. Independent coaches, specialists, and anyone whose best evidence is on their own website or a professional register.
- Roundup article
- Directory listing
- Review platform
- Professional body or registry
- A coach's own website
- Not read in this run
Same scale, different fill. The quick run uses 8 slots and every one of them is a roundup, a directory or a review platform. The long run uses 68, and 23 of those are an individual coach's own website, with another 12 on a professional register. Those two kinds do not appear in the short run at all.
Those numbers are illustrative and the shape is the point. A short read is dominated by pages whose job is to name several businesses at once, because that is the efficient thing to read when you have room for five pages. Once there is room for seventy, pages that settle one detail about one business become worth reading, and that is the first time your own site is in the running.
The honest limit: a long run reading your website is not the same as a long run recommending you. It is the difference between being in the room and being chosen, and it is covered in how Gemini decides which businesses to recommend.
A quick answer and a deep run, side by side
The two are different products of the same underlying system. Sorting them by what each one does is more useful than arguing about which is better.
| What happens | A quick answer | A deep research run |
|---|---|---|
| How long it takes | Seconds | Minutes |
| How many sources it reads | A handful | Up to hundreds, per Google |
| Whether it plans first | No | Yes, and you can edit the plan |
| What the buyer gets back | A paragraph and a few names | A multi-page report with sections |
| Which pages tend to get used | Roundups and directories | Those plus registers and individual sites |
| What gets you in | Being named by somebody popular | Being specific and confirmable anywhere |
The last row is the one worth acting on. A quick answer rewards other people writing about you. A long run also rewards you writing clearly about yourself, as long as something independent backs it up. The same split shows up in how Perplexity picks what to cite, which is covered in how Perplexity chooses its sources.
What to do about it this week
Nothing exotic. Write down the six questions a buyer would have to settle before hiring you, then make sure a page of yours answers each one in plain sentences.
The reason that works is not a Gemini trick. Google Search Central's page on AI features in Search sets the mechanical floor for anything Google-powered: a page has to be indexed and eligible to show a snippet before it can be used in an AI Overview or AI Mode answer. A long research run sits on top of the same plumbing, so the basics are not optional and a longer read does not rescue a page nobody can quote.
On what to write, the generative engine optimization study from Princeton, Georgia Tech, the Allen Institute for AI and IIT Delhi is the one piece of published measurement worth knowing. It tested changes to source pages against a benchmark and found that adding citations, quotations and statistics moved how often a source was used, while keyword stuffing did not. That is a list of things to do, made before anyone was selling a service around it.
The damaging admission: if an assistant has never heard of you, a longer read finds more of the same nothing. Depth helps a business with a thin footprint and real substance. It does not manufacture a footprint, and no amount of writing on your own site substitutes for one other place confirming you exist.
If you want to know which of those two you are, start with the free check. It asks ChatGPT 25 real buyer questions from your niche at $0 and emails a plain-English readout: your score, who got named in your place, and the one fix to start with. It runs as soon as you ask for it, and the report lands within the hour between 7am and 5pm Central, or by 10am the next morning outside those hours.
Questions we hear the most
Does Gemini Deep Research change who gets recommended?
It changes who gets the chance to be recommended. A long run reads far more sources than a quick answer does, so smaller and more specific businesses reach the point of being read at all. Being read is not the same as being chosen.
What is Gemini Deep Research?
It is a mode in the Gemini app that plans a piece of research, searches and browses on your behalf, and writes a multi-page report. Google describes it as browsing up to hundreds of websites for one question, which takes minutes rather than seconds.
Do I need to do anything different to show up in it?
No, and there is nothing to configure. The same things that make a page usable in a quick answer make it usable in a long one: plain wording, specific facts, and at least one source other than you that says the same thing.
Why do roundup articles win the fast answers?
Because one roundup names several businesses in one place, which is the cheapest way to answer a broad question quickly. A longer run has time to go past the roundup to the businesses it named, and to the ones it left out.
Is a mention inside a long report worth anything?
It is worth something when the report is what your buyer reads. A named mention with a link in a report someone commissioned about your niche is closer to a referral than a passing appearance in a list.
Does this only apply to Gemini?
No. ChatGPT, Claude and Perplexity all have longer research modes that work in broadly the same way: plan, search, browse, then write. Gemini is the one that shipped the category first, in December 2024.
How do I find out whether AI names me at all today?
Run the free check. It asks ChatGPT 25 real buyer questions from your niche at $0 and emails a plain-English readout: your score, who got named in your place, and the one fix to start with. It runs as soon as you ask for it.
Does the $500 audit look at more than one AI tool?
Yes. The $500 audit runs 125 checks across ChatGPT, Claude, Gemini, Perplexity and Google AI Overviews, and ends with the page that wins your worst question, specified section by section. It lands by this time tomorrow.
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 →