AI discoverability
When a customer asks Perplexity “best bakery in Ghatkopar” or Gemini “where can I order custom chocolate cakes near me”, the AI doesn’t read every storefront on the internet — it reads the ones it can confidently parse. This guide explains how Active Commerce makes your store readable to AI search engines and what you can do to strengthen that signal.
If you’d rather see your own store’s current state, open Active Commerce → Discoverability in the dashboard. The Schema Preview card shows the exact JSON-LD an AI crawler ingests from your store, with red chips on any field that would strengthen the schema.
Why structured data matters more for AI than for SEO
Classical search engines (Google, Bing) crawl the web and rank pages by hundreds of signals — backlinks, freshness, on-page text, schema markers. Generative-AI search engines (Perplexity, Gemini, ChatGPT, Google AI Overviews) are different: they read a smaller set of trusted sources and synthesise answers. To be cited in an AI answer, a brand needs to be unambiguous to the model in three respects:
- Entity — who is this business?
- Place — where can a customer transact?
- Offering — what category, cuisine, or service does it belong to?
schema.org JSON-LD answers all three in a single block. Without it, an AI crawler falls back to reading your H1 and meta tags — which is fine for “the brand is called X” but breaks down for “the brand is a bakery in Ghatkopar that serves Belgian chocolate cake.”
What we emit out of the box
Every Active Commerce store automatically emits structured data at three URL layers:
| URL layer | What we emit | Purpose |
|---|---|---|
{brand}.actii.me/ | Organization (multi-outlet brands) or LocalBusiness + Organization (single-outlet) | Brand entity anchor — name, logo, social profile URLs |
{brand}.actii.me/{outlet}/ | LocalBusiness (vertical subtype like Bakery, Restaurant, CafeOrCoffeeShop) | Per-outlet place — address, geo, hours, rating, phone |
{brand}.actii.me/p/{sku} | Product with Offer + AggregateRating | Per-product detail for Shopping + Merchant Center |
The schema vocabulary derives from your store’s vertical. If your brand is set up as food_beverage.bakery, the outlet-tier schema uses @type: Bakery (a more specific subtype of LocalBusiness) and emits servesCuisine: "Bakery, Cakes, Desserts". Restaurants, cafés, ice-cream shops, grocery stores, clothing stores, beauty salons, gyms — each gets the right schema subtype.
The four signals you can strengthen
The Discoverability tab in the dashboard surfaces eight AI ingestion signals; four of them are 100% in your hands.
1. PostalAddress + GeoCoordinates
The single most important field for “best X near me” queries. Active Commerce pulls your address from Google Business Profile when you connect one, or from the address you typed during onboarding. When latitude + longitude are both present, the schema emits GeoCoordinates, which unlocks map placement in AI answers and Apple Intelligence.
If the Geo coordinates row in the AI crawler checklist is red, connect Google Business Profile (one-time, two minutes) or set the location manually on the Storefront tab.
2. AggregateRating
AggregateRating (number of stars + review count) signals trust to AI models. Active Commerce reads this from your Google Business Profile cache — so it shows up automatically once GBP is connected.
We do NOT publish ratings without backing data — schema validators reject AggregateRating with only ratingValue and no reviewCount, and AI crawlers de-prioritise schema that fails validation.
3. openingHoursSpecification
AI search engines preferentially recommend businesses that are “open now.” Your structured operating hours feed openingHoursSpecification automatically — both the GBP weekday_hours shape (“Monday: 9:00 AM – 5:00 PM”) and our own store_hours map are normalised to ISO time.
4. sameAs (social profile URLs)
The sameAs field links your storefront entity to your Instagram / Facebook / YouTube / X profiles, strengthening the entity graph AI crawlers use to cross-reference brands. Add these on the Storefront tab — they’re a one-time edit with outsized impact.
AI-crawler allowlisting
The cashier-portal serves a robots.txt that explicitly allows every major generative-AI crawler:
- GPTBot + OAI-SearchBot + ChatGPT-User — OpenAI / ChatGPT
- ClaudeBot + anthropic-ai — Anthropic / Claude
- PerplexityBot + Perplexity-User — Perplexity
- Google-Extended — Gemini training + Google AI Overviews
- Applebot-Extended — Apple Intelligence
- CCBot — Common Crawl (powers many open-source LLMs)
- Bytespider — ByteDance / Doubao
You do not need to do anything to enable this — it’s on by default for every Active Commerce store. If you have an external SEO team that wants to disallow specific AI bots, contact support; we don’t support per-tenant overrides yet because the default-allow stance is what makes the storefront citeable in AI answers in the first place.
Tokenised single-use links (bill receipts at /b/{code}, payment links at /pay/{code}) are always Disallow — they’re customer-specific URLs that should never appear in a search result.
The co-occurrence play (review keyword strategy)
LLMs learn brand → intent connections from co-occurrences across reviews, blogs, and social posts. The highest-leverage source of co-occurrences you OWN is your own review-prompt copy. When you nudge customers to mention the location word and the vertical word in their Google reviews, the model strengthens the embedding-space connection between your brand and the intent customers search for.
The Review keyword opportunities card on the Discoverability tab suggests the top three terms customers should mention — typically your city, your vertical noun, and a flagship product. Deep-link to Active Reviews to wire these into your review-prompt templates.
A future release will let the SMS / WhatsApp review prompt automatically render {location_short} and {vertical_keyword} merge tags — until then, the suggestion is operator-driven copy.
Verifying your AI visibility
The fastest validation is a search-grounded AI prompt:
What are some good places to order a premium chocolate cake in Ghatkopar, Mumbai? Provide website links.
Run that query in Perplexity or Gemini with web search enabled. If your storefront shows up as a citation with {slug}.actii.me, the schema is working as designed. If not, three things to check on the Discoverability tab:
- AI crawler checklist — every row green?
- Schema preview — fields populated, no red chips?
- Brand age — newly launched stores can take 7–21 days to enter the AI corpora. Be patient.
What’s next
- Commerce WebView — how the storefront works
- Product catalog — schema for individual products
- Custom domains — when to use a custom domain vs. the actii.me subdomain