The full-cycle platform for AI search optimization.

Chaice instruments every answer engine and closes the loop — controlled prompt runs, scored deltas, drafted fixes, published content, re-measured results.
Dashboard
Prompts
Studio
Competitors
Sources
Topics
Integrations
Integrations/ GSC · Shopify · CMS
Integrationsyour data in, your content out
Google Search Consoleconnected ✓
Shopifyconnected ✓
WordPressconnect
Contentfulconnect
GA4connect
Importing1,204 pages · 8 prompts seeded from your search data
Point Chaice at your stack — search data in, content out.
Run your baseline
100 free test creditsNo card required
Checking your prompts every day
ChatGPTGoogle AI OverviewsPerplexitysoonGeminisoonCopilotsoonGroksoonClaudesoonChatGPTGoogle AI OverviewsPerplexitysoonGeminisoonCopilotsoonGroksoonClaudesoon
Protocol 01 · Observe

Know exactly where you stand.

One score across every engine, and your slice of the category answer measured against every brand AI names.

  • Combined and per-engine visibility
  • Competitor lines on the same axis
  • Share of voice against all 47 brands
Dashboard/ Stumpytown
AI visibility
11/100
20 this month
All enginesChatGPTAI Overviews
013253850JUL 13JUL 27
Share of voice
5%
#4of 47
Blue Bottle Coffee 6%Trade Coffee 5%Bean Box 5%You · Stumpytown 5%La Colombe 5%Starbucks 5%
Protocol 02 · Trace

Know why competitors win.

AI doesn't invent an answer, it reads pages. Chaice traces every prompt back to the sources behind it, and shows which of them mention you.

  • Citation flow from prompt to source
  • Which cited pages already mention you
  • The widest source that ignores you entirely
Sources/ Stumpytown
Sourceshow answers on your prompts flow through sources · ribbon = citations3%of cited pages mention you
Your 8promptsreddit.com · 5 promptsfacebook.com · 3 promptsseriouseats.com · 3 promptswired.com · 3 promptsallrecipes.com · 2 prompts · cites youamazon.com · 2 prompts
One violet ribbon carries you (allrecipes). The widest — reddit.com, cited on 5 of 8 prompts — flows past you entirely: your highest-leverage pitch.
Protocol 03 · Prioritise

Know what to fix first.

Every attribute AI judges your category on, plotted by the weight it carries against the credit AI gives you. Top-right is your moat. Bottom-right is the work.

  • Attribute weight × credit, scored daily
  • Fix-first, moat, quiet strength, let go
  • A written read on the biggest single gap
Studio/ Stumpytown
What AI prioritizesweight AI gives each attribute × how much AI credits you
Weight AI gives the attribute →
Evaluation and measurement carries the heaviest weight on your prompts and AI credits you 0/100 for it — the definition of fix-first.
Protocol 04 · Synthesise

Refine it in the Studio. Publish. Get ranked.

DraftRefinePublishRe-measure

  • Drafts grounded in the pages AI cites
  • Marked insertions — you approve every line
  • Prompts re-run after publish to score the lift
Studio/ draft · cold brew, measured
targets your fix-first gap
Products
Original Cold Brew · 32 oz86Concentrate · 64 oz71Nitro Cans · 6-pack62Single-Origin Beans · 12 oz78Decaf Cold Brew · 32 oz55
Original Cold Brew · 32 ozPreview‹v4›
Title
Cold brew, measured: how we test strength and smoothness
56 charactersfits AI answer snippets
Description · en-US

Every batch we ship is brewed from the recipe we publish here — ratio, steep time and filtration stated plainly, so anyone can check the claims.

We steep at 1:8 for 18 hours and log total dissolved solids for every batch — a measured strength readout answer engines can cite.

Serving guide: neat, over ice, or 1:1 with milk — the versatility question engines ask most.

370 characters2 of 2 AI insertions
Save draftPublished ✓
AI readiness
86/100
+52 vs current page
Target prompts · this product
“best coldbrew coffee”2.4k/mo“cold brew vs iced coffee”1.1k/mo
What AI prioritizes
attributes weighted from your tracked answers
Evaluation and measurementadded
heaviest weight · no brand credited yet
Drinking versatilityadded
credited today:— not you
Ranking moved: #7 → #4 on “best coldbrew coffee” · next read tomorrow 06:00
Protocol 05 · Examine

Every prompt, opened up.

Drill from the dashboard into a single question and see the whole field: who's named, by which engine, in what terms, and whether your domain was cited at all.

  • The full ranking per question
  • Sentiment and owned attributes per brand
  • Mentioned-but-never-cited, flagged
Prompts/ “best coldbrew coffee”United States
Your share of voice
10%
All enginesChatGPTAI Overviews
Jul 14 — now
The field · visibility
La Colombe leads at 20% · all 8 in ranking

Mentioned, never cited

AI Overviews and ChatGPT name you without citing your domain.

PositiveFlavor qualityAvailabilityyou own
Linked content · 3Original Cold BrewproductConcentrateproductBrewing guidepage · new
Ranking8 brands in the field · “best coldbrew coffee”
#BrandMentioned bySentimentWhat AI prioritizesSourcesShare of voice
1La Colombe Coffee Workshoplacolombe.comFlavor qualitySmoothness+2reddit.comseriouseats.comsprudge.comCited sourcesreddit.comseriouseats.comsprudge.com
2Wandering Bear Coffeewanderingbearcoffee.comCaffeine strengthFlavor quality+1reddit.comwirecutter.comfoodandwine.comCited sourcesreddit.comwirecutter.comfoodandwine.com
3Califia Farmscalifiafarms.comSmoothnessVersatilityreddit.comseriouseats.combonappetit.com+6Cited sourcesreddit.comseriouseats.combonappetit.comfoodandwine.comtasteofhome.comepicurious.comthespruceeats.comdelish.comeater.com
4Dunkin'dunkindonuts.comFlavor qualityValue for money+1reddit.comCited sourcesreddit.com
5Starbucks Coffee Companystarbucks.comFlavor qualityVersatilityeater.comCited sourceseater.com
6Stokstok.comFlavor qualitySmoothness+1tasteofhome.comCited sourcestasteofhome.com
7STStumpytown · younot citedFlavor qualityAvailability+10
0brands compete in a single category answer
0%of cited pages mention you today
0answer engines checked, every day
0of those cited pages are yours

“We could see the traffic falling. Chaice was the first thing that told us where it was going.”

NBHead of Growth, DTC coffee brandplaceholder — swap for a real customer quote

Straight answers.

ChatGPT and Google AI Overviews today. Every prompt runs on both, so you can tell an AI Overviews collapse from a ChatGPT one. Perplexity, Gemini, Copilot, Grok and Claude are next — we list an engine as live only once we actually query it.

It combines whether you're named on each tracked prompt, whether your domain is cited as a source, and how prominently that citation sits among the sources the answer used — normalised to 0–100 so it's comparable week to week and against competitors. Sentiment is measured and reported separately rather than folded into the score.

Daily by default. Answer engines are non-deterministic, so a single check is noise — the score you see is a rolling read across repeated runs, which is why the trend matters more than any one day.

Chaice tells you what to change and drafts the content. Publishing stays with you — nothing goes live without an explicit action.

Related, but not the same. Ranking on a results page and being named inside a generated answer are different outcomes with different inputs — AI leans heavily on third-party sources like Reddit, review sites and listicles that classic SEO tooling never measured.

Walk into your next meeting with the numbers.

Baseline your domain today. Every recommendation Chaice makes ships with the data behind it.

100 free credits · cancel any time