- Jul 26, 2016
- 2,273
- 3,649
Visibility means owning a topic, not a prompt.
Ranking for one prompt doesn't mean you're visible. The brands that win are the ones showing up across multiple related prompts in the same topic.
Most topics have nobody in them.
Only 15.2% of the 1,094 topics studied had a clear brand leader. So around 85% are still up for grabs, including in high demand categories.
Traditional SEO isn't enough on its own.
Broad domain strength gets you into the fight, it doesn't decide who wins the category.
It compounds once you're there.
Brands with a solid lead kept their position in 90% of month over month comparisons. Get in early and it holds.
NOW CHECK YOUR OWN CATEGORY
Everyone reads a study like this and goes straight to building. Find out where you actually stand first. Took me under an hour and cost nothing.
Step 1. Pick a topic and write four prompts.
Topic, not keyword. Air purifiers, not one search term.
Four prompts, all buying intent, each a different shape:
Step 2. Run them and write down every brand named.
New chat each time, memory off, logged out, same model. Don't send follow ups. I ran all four in one sitting.
Six brands in that one: Coway twice, Blueair twice, Rabbit Air, IQAir.
Step 3. Count them up.
The study's bar is one brand in 4 of 5 prompts with a clear lead over second. I ran four prompts, so I used 3 of 4 as the equivalent.
My run:
Coway 4 of 4
Levoit 3 of 4
Blueair 3 of 4
Winix 2 of 4
Rabbit Air 1 of 4
IQAir 1 of 4
Coway is in every single answer and one clear ahead of the pack. That category has an owner.
Worth knowing: I ran a rougher version of this a few weeks back with different prompts and Coway came third. Different prompt set, different winner. So don't treat one run as gospel, and re-run it monthly.
Step 4. Now look at where the answers are sourced from. Different scoreboard.
This is the part nobody checks, and it's where it got interesting.
The brands in the answer and the sites cited underneath are two different lists, and which sites get cited changes completely depending on how the question is phrased.
Situational prompt, the bedroom and cat allergies one:
TechGearLab, Air Purifier First, Smart Home Explorer. Small independent review sites. Sites people in here could build.
Same category, budget prompt:
Levoit, Cowaymega, one small site. Almost entirely manufacturer domains.
And the straight "best right now" prompt back in step 2 pulled Real Simple, WIRED, Consumer Reports, Forbes, TechRadar and HouseFresh. Big media, nearly all of it.
Three different question shapes on one topic, three completely different source pools.
So the citation slot isn't one competition, it's several. Generic best-of questions are getting answered from big media and you're not outranking Forbes this year. Price questions are getting answered from the manufacturers themselves. The situational questions, the ones with a room size and a pet and a sleep problem in them, are where the independent review sites are actually getting cited.
That's the gap. It's also the kind of content most people skip because it doesn't have search volume attached to it.
Step 5. Read your result.
Nobody at 3 of 4 or better: category is open, worth building for.
One brand at 4 of 4 with clear daylight, like mine: they own it, and the study says leaders hold position 90% of the time month to month. Expect a fight on the brand layer.
Answers coming back with no brands at all: you've written informational prompts, or it's a topic with no commercial slot in it.
Now look at your source columns separately from your brand column. Which question shapes in your topic are pulling independent sites rather than big media or manufacturers? Those are the ones worth writing for.
WHAT TO DO ONCE YOU KNOW
Build comprehensive content around the topic rather than scattered one off pages.
Earn mentions in the places these models actually pull from. Reddit, review sites, industry publications, expert commentary.
Strengthen entity signals so your brand is consistently associated with the topic.
And re-run your four prompts monthly. Mine moved in a few weeks. It's the only way to know if any of it is working.
AI search isn't about winning individual queries. It's about becoming the brand the model consistently associates with an entire category, on the question shapes where there's actually a seat available.
Post your numbers if you run it. Interested whether most topics really do have no clear leader, especially in the niches people in here actually work in.
Ranking for one prompt doesn't mean you're visible. The brands that win are the ones showing up across multiple related prompts in the same topic.
Most topics have nobody in them.
Only 15.2% of the 1,094 topics studied had a clear brand leader. So around 85% are still up for grabs, including in high demand categories.
Traditional SEO isn't enough on its own.
Broad domain strength gets you into the fight, it doesn't decide who wins the category.
It compounds once you're there.
Brands with a solid lead kept their position in 90% of month over month comparisons. Get in early and it holds.
NOW CHECK YOUR OWN CATEGORY
Everyone reads a study like this and goes straight to building. Find out where you actually stand first. Took me under an hour and cost nothing.
Step 1. Pick a topic and write four prompts.
Topic, not keyword. Air purifiers, not one search term.
Four prompts, all buying intent, each a different shape:
- Straight recommendation: what's the best air purifier right now?
- Budget: what air purifier should I buy under $300?
- Situation: I have a 400 sq ft bedroom, cat allergies and I'm a light sleeper. What air purifier should I get?
- Replacement: my air purifier died after two years, what should I replace it with?
Step 2. Run them and write down every brand named.
New chat each time, memory off, logged out, same model. Don't send follow ups. I ran all four in one sitting.
Six brands in that one: Coway twice, Blueair twice, Rabbit Air, IQAir.
Step 3. Count them up.
The study's bar is one brand in 4 of 5 prompts with a clear lead over second. I ran four prompts, so I used 3 of 4 as the equivalent.
My run:
Coway 4 of 4
Levoit 3 of 4
Blueair 3 of 4
Winix 2 of 4
Rabbit Air 1 of 4
IQAir 1 of 4
Coway is in every single answer and one clear ahead of the pack. That category has an owner.
Worth knowing: I ran a rougher version of this a few weeks back with different prompts and Coway came third. Different prompt set, different winner. So don't treat one run as gospel, and re-run it monthly.
Step 4. Now look at where the answers are sourced from. Different scoreboard.
This is the part nobody checks, and it's where it got interesting.
The brands in the answer and the sites cited underneath are two different lists, and which sites get cited changes completely depending on how the question is phrased.
Situational prompt, the bedroom and cat allergies one:
TechGearLab, Air Purifier First, Smart Home Explorer. Small independent review sites. Sites people in here could build.
Same category, budget prompt:
Levoit, Cowaymega, one small site. Almost entirely manufacturer domains.
And the straight "best right now" prompt back in step 2 pulled Real Simple, WIRED, Consumer Reports, Forbes, TechRadar and HouseFresh. Big media, nearly all of it.
Three different question shapes on one topic, three completely different source pools.
So the citation slot isn't one competition, it's several. Generic best-of questions are getting answered from big media and you're not outranking Forbes this year. Price questions are getting answered from the manufacturers themselves. The situational questions, the ones with a room size and a pet and a sleep problem in them, are where the independent review sites are actually getting cited.
That's the gap. It's also the kind of content most people skip because it doesn't have search volume attached to it.
Step 5. Read your result.
Nobody at 3 of 4 or better: category is open, worth building for.
One brand at 4 of 4 with clear daylight, like mine: they own it, and the study says leaders hold position 90% of the time month to month. Expect a fight on the brand layer.
Answers coming back with no brands at all: you've written informational prompts, or it's a topic with no commercial slot in it.
Now look at your source columns separately from your brand column. Which question shapes in your topic are pulling independent sites rather than big media or manufacturers? Those are the ones worth writing for.
WHAT TO DO ONCE YOU KNOW
Build comprehensive content around the topic rather than scattered one off pages.
Earn mentions in the places these models actually pull from. Reddit, review sites, industry publications, expert commentary.
Strengthen entity signals so your brand is consistently associated with the topic.
And re-run your four prompts monthly. Mine moved in a few weeks. It's the only way to know if any of it is working.
AI search isn't about winning individual queries. It's about becoming the brand the model consistently associates with an entire category, on the question shapes where there's actually a seat available.
Post your numbers if you run it. Interested whether most topics really do have no clear leader, especially in the niches people in here actually work in.