AI visibility report
Nike
nike.com · 1 Aug 2026
10 buyer questions asked across Google AI Overview, ChatGPT, Google AI Mode, Google Gemini, producing 40 answers. Every figure below is a count of those answers.
AI Visibility Analysis
How visible is the brand across AI engine answers, and who is being recommended instead?
GEO score
18/100How visible the brand is across the AI engines, how often they recommend it, and how favourably they describe it.
AI visibility
21/100Share of voice, averaged across the engines. On each engine: how many of that engine's brand mentions were yours, counting each brand once per answer. Every tracked question counts, including the ones naming your own brand — the split between the two is reported separately rather than removed. Presence and citation rate are measured and reported beside this, but not scored.
Prompt coverage
70%7 of 10 questions name the brand
Share of voice
21%Of every brand named
Citation rate
15%Answers that read a page on nike.com
Share of voice, by engine
21%
average across 4 engines
- ChatGPT20% · named in 6 of 10
- Google Gemini20% · named in 6 of 10
- Google AI Overview23% · named in 7 of 10
- Google AI Mode23% · named in 7 of 10
Funnel vulnerability, by buyer intent
0/8 on problem-solving questions — rivals are intercepting buyers before brand selection begins.
- Problem-solving0 of 8 named
- Buying decision4 of 8 named
- Comparison7 of 8 named · 3 cited
- Recommendation15 of 16 named · 3 cited
Nike is named in 15 of 16 recommendation questions and 0 of 8 problem-solving ones.
Independent sources
78%Cited pages nobody in this market owns
Rival-controlled
16%Cited pages a tracked rival owns
Questions lost outright
4A rival named where the brand is not
Own citation share
5%Of the citation pool, the brand's own pages
Source types feeding the answers
- Editorial & review media
- 46%17
- Community & forums
- 19%7
- Rival official sites
- 16%6
- Retail & marketplaces
- 14%5
- The brand's own site
- 5%2
Rivals taking your prompts
- Brooks2 answers
- Asics2 answers
- Hoka2 answers
- Adidas1 answer
- On Running1 answer
Top cited sources & domains
5% of them are nike.com. The rest are pages Nike does not control.
- reddit.comCommunity6Earned, not bought
- runnersworld.comEditorial4Needs a pitch
- runrepeat.comEditorial4Needs a pitch
- fleetfeet.comRetail3Fill in a form
- brooksrunning.comBrooks2Earned, not bought
- healthline.comEditorial2Needs a pitch
- nike.comNike2
Sentiment Analysis
Is AI recommending the brand, and is it saying positive things?
Average brand position
2.5
2nd of 7 brands named
- Brooks1.7
- Nikeyou2.5
- New Balance2.8
- Hoka2.8
- Adidas2.9
- Asics3
- On Running3.8
named first (1.7)named last (3.8)
Sentiment & word breakdown
- 4unfavourable
- 12neutral
- 10favourable
Average 56 out of 100, where 50 is neutral — neutral. 4 answers read unfavourably — worth opening before a client does.
Hallucination & contradiction alert
- Nike's ZoomX foam is not intended for daily training mileage.Google AI Overview
Nike's ZoomX, used in the racing models, wears out much faster by design and is not meant for daily mileage.
Your site says: The Invincible 3 is described on nike.com as a ZoomX daily trainer built for everyday mileage. Asked as “Nike or Adidas running shoes — which last longer?”.
- Nike's marathon racing shoes cost around $250 to $300.Google AI Overview
All use a carbon plate with a highly resilient foam, all cost around $250 to $300, and all — the Nike models included — have a short racing lifespan, so most runners save them for race day.
Your site says: Alphafly 3 is listed at $285 and Vaporfly 4 at $270 on nike.com. Asked as “What are the best marathon racing shoes right now?”.
- The Nike Invincible is less stable than rival max-cushioned shoes at higher body weights.Google AI Overview
The Nike Invincible offers similar ZoomX cushioning and a very soft ride, though it feels less stable than the others at higher body weights.
Your site does not say this anywhere it was read. Asked as “Best cushioned running shoes for heavier runners?”.
Checked against nike.com · 2 assertions are not addressed on the site at all.
Action Plan
What exact tasks should the team execute tomorrow morning?
Action plan hub
Open the action plan →29 pages you need to get into
Each one feeds two or more answers in this run and mentions you on none of them. Between them they hold 10 of your questions. A listing, a pitch or a submission. Nothing on your site moves it.
See all 29 →6 questions need a page of your own
You are named in these answers and somebody else's page is cited. Written and published on your own domain.
See all 8 →6 questions a rival's own page holds
5 are assembled in part from a page a competitor writes and controls, and 1 more names you after them. An existing page, restructured to answer first.
See all 6 →
Closest to being won
- +2.5points of GEO score
Where can I buy running shoes with the best return policy?Answered without you
Named in 0 of 4 answers, cited in 0. Take the ground: nothing on the domain was close enough to bring in.
- +2.5points of GEO score
Which running shoes help with knee pain?Answered without you
Named in 0 of 4 answers, cited in 0. Hoka, Brooks and Asics hold it. Take the ground: nothing on the domain was close enough to bring in.
- +2.5points of GEO score
How do I know when to replace my running shoes?Answered without you
Named in 0 of 4 answers, cited in 0. Brooks hold it. Take the ground: nothing on the domain was close enough to bring in.
- +0.6points of GEO score
What are the best marathon racing shoes right now?Named on some engines
Named in 3 of 4 answers, cited in 3. Adidas, Asics and On Running hold it. Cover the question on a page of its own, not inside another.
- Ground already held
What are the best running shoes for flat feet?Behind rivals
Named in 4 of 4 answers, cited in 0. Get onto one of the pages the ordering is being read from.
Already named on every engine — winning the citation lifts your citation rate 10 points but not the GEO score
What this report does not cover
- One market and one language. Answers vary by country, and nothing here measures that.
- Each question was asked on each engine on this run. An answer is a sample, not a constant.
- Whether a fix was carried out is not recorded anywhere in this product — the next run is what reports the effect.
- Paid placement, and anything an engine shows outside its written answer.