Answer Engine Optimization

Target the intent. Claim the response.
NBA · Phiture 2026 · GEO / AEO experience and what we found
cover
What we will cover

How the NBA app earns the recommendation.

We ran the study on the NBA app before this meeting, so everything here is measured, not theoretical.
01
The shift in app discovery
02
What the AI recommends for the NBA app today
03
The two surfaces that determine visibility
04
Our approach to Answer Engine Optimization
05
Measurement and what a pilot looks like
02
01 · the shift

Discovery is moving from search to the ask.

How it is changing
Before: “basketball scores”
Now: “how do I watch my team live tonight without cable, and get alerts?”
Fans ask ChatGPT and Gemini outside the store, and now, through Ask Play, inside it too. The job is no longer only to rank. It is to be the app the AI names.
Volumes are still small, so this is not yet a big traffic driver. But the surfaces and consumer adoption are changing fast, and the advantage goes to whoever is the named answer before the volume arrives.
The scale of the shift
20%
of informational queries now happen on chatbots1
34%
of Gen-Z use chatbots for research2
5.5×
faster search growth than Google3
+31%
conversion rate vs non-branded organic search4
11×
YoY growth in ChatGPT + Gemini app referrals, across some of our other clients
1Digital Applied   2Semrush   3Bond Cap   4Search Engine Land 2026. Referral growth from Phiture client data.
03
02 · what the AI recommends today

It owns the league. It loses the fan.

NBA app
watching the game
Live games
C100 G100
Without cable
C100 G100
Highlights
C100 G83
Out of market
C100 G100
On the TV
C100 G100
Free
C100 G83
Listen
C100 G100
Catch up
C100 G100
NBA app
following the game
Live scores
C100 G67
Stats
C100 G17
Standings
C100 G100
News
C100 G100
Team alerts
C100 G67
Follow a player
C33 G33
Fantasy
C0 G0
Fan chat
C50 G0
named 60%+   30–59%   under 30%    C = ChatGPT, G = Gemini Share of Answer; cell colour is the mean of the two. One app, two job families.
Anything about the league is won: every watching job, plus news, standings and scores. Anything personal to the fan is lost: your fantasy team (0% on both engines), the player you follow (33%), and the conversation with other fans. Those go to Sleeper, Yahoo Sports, theScore and Reddit.
100%
watching the game
ChatGPT, and 96% on Gemini
73%
following the game
ChatGPT, and 48% on Gemini
One app, 16 unbranded jobs, two engines. Prompts say “basketball”, never “NBA”.
192 grounded runs: 16 unbranded clusters × 2 phrasings × 3 reps × 2 engines (gpt-5 and gemini-3.5-flash, both with web search), US, Sep 2026. “NBA” is the league name as well as the brand, so a league-name prompt would be near-tautological; these prompts use the category word instead.
04
02 · where the answer goes elsewhere

The gaps are personal, and the listing never claims them.

GPT
Gem
Stats and box scores
100%
17%
vs Sofascore, Yahoo Sports, theScore
Fantasy and picks
0%
0%
vs Sleeper, Yahoo Sports, ESPN
Follow a player
33%
33%
vs Yahoo Sports, theScore, Bleacher Report
Talk with other fans
50%
0%
vs Bleacher Report, Reddit, Discord
Live scores
100%
67%
vs FlashScore, Sofascore, Yahoo Sports
The weakest “following” jobs, and who is named instead. Where the two engines disagree, as on stats, the picture is softer than a single engine suggests.
Now read the store listing. It says live nine times, free nine times and stream eight times. It is written as a broadcaster.
The words fans actually ask in are absent. “cable”, “replay”, “radio”, “listen”, “alerts”, “notifications” and “fantasy” appear zero times in the description. The engines cannot quote what the listing never says.
This is the cheapest fix in the deck. The app already does most of these jobs. The listing, and the pages the engines cite, simply never claim them.
Listing text from the live US Google Play listing for com.nbaimd.gametime.nba2011, Sep 2026. Shares from the same study.
05
03 · the two surfaces

AI adds two layers on top of your listing.

Layer 1 · above the store
Win the answer
Be the app ChatGPT and Gemini name, via the store listing, NBA-owned pages and the sources they cite. Answer Engine Optimization.
Layer 2 · inside the store
Win the store
Be the app Ask Play recommends when a fan describes the job in the store, by strengthening the listing and quality signals its ranking actually uses.
Measure: Phiture’s rig tracks Share of Answer in ChatGPT and Gemini above the store, and probes Ask Play on a real device inside it. We also checked Play’s AI collections across 12 unbranded basketball queries; none surfaced at all, so that surface is not a lever here yet.
06
03 · how Ask Play picks

What decides who Ask Play recommends.

We got the store assistant to explain its own selection. It filters first, then ranks whatever survives. That order matters more than the ranking itself.
First it cuts
Anything rated under 4.0 stars.
Anything that does not look actively maintained.
Anything whose listing does not explicitly support the job.
Then it ranks on
User rating, the heaviest single factor.
Download volume, and editorial picks like Editors’ Choice.
How recently you updated.
The NBA app clears the gates comfortably: 4.7★ on iOS, 4.5★ on Google Play, 10m+ installs and monthly releases. So the lever here is not quality, it is explicit job support: the listing has to say the job, in the words fans ask it in.
Ask Play asked to explain its own selection, physical Android device, US, Sep 2026. Self-reported rather than verified internals, and it describes itself differently in other categories, so treat the shape as the signal. Ratings from the live US store listings. A device probe on basketball queries is the immediate next step.
07
04 · our approach

We speak the language of both ASO and AI.

ASO optimises for the search
Your proven foundation
Metadata: titles, short and long descriptions.
Keyword strategy for high-intent terms.
Listing structure built for conversion.
Rank, conversion and install tracking.
AEO optimises for the ask
The performance edge
Answer formatting: turn long descriptions into clear answers.
Cluster strategy: align content with the jobs fans ask.
Cited-page alignment: the off-listing sources the engines quote.
Share of Answer testing: measure what actually moves.
The store listing is the single largest citation source ChatGPT uses for app recommendations, 47.5% of citations. Restructuring it around the job is the one change that lands on ChatGPT, Gemini and Ask Play at once.
AppTweak analysis of 125K+ ChatGPT responses across 9,489 app prompts, US, May 2026.
08
04 · how we win

How we win a category job, step by step.

1 · Prioritize
Pick the unbranded jobs the NBA app should own, by fit, demand and value.
2 · Fix the listing
Say the job in the words fans ask it in, across title, subtitle, description and screenshots.
3 · Cited pages
NBA.com first, then the third-party pages the AI already trusts.
4 · Measure
Did Share of Answer move? Next job.
Start where the app already delivers but the answer goes elsewhere: following a player, team alerts and fan conversation. These are listing and citation gaps, not product gaps, which makes them the fastest wins available. Fantasy is the one real product question.
09
05 · measurement

How we measure AEO impact: one KPI funnel.

Read left to right, vanity to sanity. Each stage has its own live view, so one optimisation shows up as movement all the way to installs.
1 · Awareness
Share of Answer
0255075100W1W2W3W4W5W6
Cluster visibility over time. Illustrative; latest point = real study.
2 · Intent
AI referral clicks
W1W2W3W4W5W6
Clicks from ChatGPT & Gemini through to the store listing (App Store Connect).
3 · Action
App installs
04008001200AprMayJunJulAugSep
ChatGPT + Gemini installs, from App Store Connect App Referrer.
Caveat: counts only users shown a direct listing citation who tap it, not those who then search the store by name, so it under-reports. A “How did you hear about us?” in-app prompt uncovers the fuller picture.
Illustrative dashboard views; the latest cluster reading is the real study. KPI-funnel framework (Precis × Phiture); installs from App Store Connect App Referrer once connected.
10
05 · what a pilot looks like

A 90-day pilot, on the jobs that matter most.

One app, iOS and Google Play, one market, two experiments per platform.
Phase 1 · weeks 1–4
Set up & audit
Share of Answer baselined at day zero in StorePulse, across ChatGPT and Gemini.
Identify which pages the engines actually cite.
Probe Ask Play on device for basketball jobs.
Phase 2 · weeks 5–8
Implement
Rewrite the listing around the following jobs, in the words fans ask them in.
Align the cited off-listing pages.
One listing experiment per platform.
Phase 3 · weeks 9–12
Measure & expand
Report Share of Answer, recommendation rate and accuracy vs baseline.
Tie to AI referral clicks and installs.
Plan the next market and cluster.
Fastest first win: following a player and team alerts. The app already does both; the listing never claims them, so this is a positioning and citation fix rather than a product one.
11

Let’s discuss.

close