Starter
$99/ mo
- Up to 500 products tracked
- Weekly refresh, six engines
Product-level AI visibility · ecommerce AEO
The AI visibility platform that sees every product.
Your buyers ask ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude, and Copilot what to buy before they reach your store. eCommerce Insights tracks which of your products those answers cite, scores every PDP twice — citation and agent-readability — and ships the title, schema, and metafield fixes as diffs a human approves.
How it works
Four steps from connected catalog to shipped fix.
No onboarding project, no services engagement. Connect, scan, read the two scores, approve the diffs.
Step 01
Connect a channel
Connect Shopify through the admin API, or link Amazon, Adobe Commerce, BigCommerce, WooCommerce, or a headless storefront. Catalog, variants, and metafields are read in under five minutes. No developer required.
Step 02
Scan the catalog
Every SKU runs against a category-tuned prompt bank across ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude, and Copilot. First results within 24 hours; weekly refresh on Starter, daily on Growth.
Step 03
Read two scores per product
Each SKU gets a citation score (is the engine recommending it) and an agent-readability score (can a shopping agent parse the PDP). Every sub-factor is visible, so the score is a work queue, not a verdict.
Step 04
Ship the fixes
Each failing SKU gets a reviewable diff: proposed title, description, bullets, schema, and metafield values. A human approves; eCommerce Insights pushes to Shopify Early access or exports CSV for every other channel.
Capability · Track
AI search visibility: every SKU, every engine, on your plan's cadence.
Brand-level trackers tell you the brand was mentioned. eCommerce Insights resolves every AI answer to a specific SKU, a specific prompt, and a specific engine — because "the brand showed up" doesn't tell you whether the answer recommended your best-seller or your clearance bin.
Each scan runs a rotating, category-tuned prompt bank and records which SKU was cited, which sources were cited alongside it, and how the engine phrased the recommendation. Prompts are generated from your collections and best-sellers, then rotated each scan, and segmented by intent: "best" behaves differently from "cheapest," "gift for," and "alternative to."
Starter refreshes weekly; Growth refreshes daily. Co-citation data shows the real competitive set per query — often review sites and Reddit threads, not the competitors in your deck. Full methodology on the SKU-level tracking page.
Prompt · "best merino base layer for cold-weather running"
| Engine | Status | Citations |
|---|---|---|
| ChatGPT | Cited | 3 |
| Perplexity | Cited | 1 |
| Google AI Overviews | Missing | 0 |
| Gemini | Missing | 0 |
| Claude | Cited | 2 |
| Copilot | Cited | 1 |
Co-cited sources: rei.com, outdoorgearlab.com, reddit.com/r/ultralight. Illustrative.
Capability · Score
Two scores per product: is it cited, and can an agent read it.
One worked example, factor by factor. MERINO-BL-MD is an illustrative SKU — a merino wool base layer, men's medium — but the factors and the failure modes are exactly what every scan reports.
Citation score — why engines aren't recommending it
| Factor | Score |
|---|---|
| Structured data completeness — Product schema fields present | 16 / 20 |
| Citation surface — citations vs catalog peers on the prompt set | 11 / 20 |
| Entity clarity — brand consistency, canonical URL, title specificity | 15 / 20 |
| Answer coverage — does the PDP answer the top buyer questions | 13 / 20 |
| Review signal — volume, recency, review schema surfacing | 16 / 20 |
Citation score71 / 100
Diagnosis: the PDP is well structured but rarely cited — the prompt set routes to review sites because the page doesn't answer "fits true to size?" or "how warm is 200gsm?"
Agent-readability score — can an agent parse the PDP
| Factor | Score |
|---|---|
| Product JSON-LD completeness — 8 of 11 recommended fields | 17 / 25 |
| robots.txt admittance — Google-Extended blocked | 15 / 20 |
| Machine-readable price and availability in offers | 16 / 20 |
| Discoverable returns and shipping policies | 10 / 20 |
| Agentic-checkout wiring (ACP/UCP feeds, where applicable) | 6 / 15 |
Agent-readability score64 / 100
Diagnosis: a shopping agent can find the price but not the returns policy, and one AI crawler is blocked outright. Both fixes ship in the same diff.
Illustrative scores for a hypothetical SKU. Real scores come from live scans. Methodology: citation score · agent-readability score · PDP Score docs.
Capability · Fix
Reviewable diffs, not silent rewrites.
For every SKU below the threshold you set, eCommerce Insights produces a diff: proposed title, description, bullet set, schema block, and metafield values, shown next to what's live today. A merchandiser reviews, edits, and approves. Nothing ships without a human decision.
On Shopify, approval can push the change straight through the admin API — including metafield writes for GTIN, material, and the rest of the Product schema, in the correct namespace. Shopify push · Early access Every push is logged, attributed to the approver, and reversible in one click.
On Amazon, the diff covers title, bullets, A+ Content, and backend keywords for Seller Central. On every other channel, and on every plan, approved diffs export as CSV. More on the workflow: PDP optimization and the listing rewrite docs.
PDP diff · MERINO-BL-MD score +14
title
- Merino Base Layer Men's Medium
+ Merino Wool Base Layer for Men, Medium
+ — 200gsm, Cold Weather
body
+ "Fits true to size" answer added
+ Care instructions (machine wash cold)
schema.Product
+ gtin: 00860009123456
+ material: "100% merino wool"
+ hasMerchantReturnPolicy: 30-day returns
metafield: custom.gtin
+ 00860009123456Illustrative diff. Real diffs are generated from the scan and your live PDP.
Channel-aware engine
Recommendations match the channel they ship to.
A Shopify metafield recommendation is useless on Amazon; a backend-keyword rewrite is useless on Shopify. The scoring engine routes every product by channel.
Shopify · primary
Six-engine scan, metafield-deep output.
Shopify SKUs run the full six-engine scan and get diffs that speak Shopify: titles, body HTML, Product JSON-LD, and metafield values in the right namespace. Variants are read and scored where engines cite specific sizes or colors. Push lands through the admin API on Growth and up. AI visibility for Shopify →
Amazon · marketplace
Rufus and COSMO scoring, listing-native output.
Amazon SKUs route to Rufus and Amazon's COSMO model, scored across its 15 commonsense intent relations — the "used for," "worn with," "bought for" links between products and buyer intent. Recommendations cover title, bullets, A+ Content, and backend keywords for Seller Central. AI visibility for Amazon →
Adobe Commerce, BigCommerce, WooCommerce, and headless storefronts get the same six-engine scan with vendor-specific recommendations. See all six channels.
Inside the platform
The feature ledger.
Every feature resolves to a SKU, a prompt, and an engine. Documentation for each is public — read it before you trial.
PDP Score
The two-score model per SKU. Channel-aware — Amazon SKUs route to Rufus/COSMO.
Read docs →
AI Agent Lens
15 AI bots checked against your robots.txt, plus the plain-text view an agent actually parses.
Read docs →
Prompt Runs
Run the same buyer prompt across engines side by side and see who gets cited.
Read docs →
Listing rewrite
Per-field diff capsules: title, body, JSON-LD additions, Shopify metafield values.
Read docs →
Image Enhancer
Feature-sentiment mapping, mobile glanceability, competitive benchmarking, creative brief.
Read docs →
Sentiment
Themes from real reviews, and the tone gap between customer sentiment and AI descriptions.
Read docs →Integrations
Works with the stack you already run.
Shopify
Shopify Plus
Amazon Seller Central
Adobe Commerce
BigCommerce
WooCommerce
GA4
Google Search Console
Agencies and holding companies
Running this across multiple brands?
Agency & Enterprise plans add multi-brand dashboards, per-client seats, and white-labeled reports (Early access). One login, every client catalog, a baseline report in week one. See for agencies and for multi-brand portfolios.
Pricing
Flat monthly. No sales gate below Agency & Enterprise.
Growth
$349/ mo
- Up to 2,500 products tracked, daily refresh
- Shopify push Early access
Agency & Enterprise
Custom
- Multi-brand, agencies
- SSO, DPA, dedicated CSM
Questions buyers ask
What AI engines does eCommerce Insights track?
Six engines on every plan: ChatGPT (including ChatGPT Shopping and Instant Checkout), Perplexity (including Perplexity Shopping and Buy with Pro), Google AI Overviews and AI Mode, Gemini, Claude, and Copilot. SKUs on a connected Amazon channel are also scored against Rufus and the COSMO intent model; Sparky (Walmart) is available on Agency & Enterprise plans.
What does the agent-readability score actually check?
Five factors per PDP: Product JSON-LD completeness, robots.txt admittance for AI crawlers, machine-readable price and availability, discoverable returns and shipping policies, and agentic-checkout wiring where applicable. Each factor is reported separately with the specific fields or bots that failed, so the fix is obvious from the score.
How is the citation score calculated?
Five weighted factors, refreshed with every scan: structured data completeness, citation surface (how often the SKU is cited on the tracked prompt set), entity clarity, answer coverage (whether the PDP answers the category's top buyer questions), and review signal. Every sub-score is visible per SKU, per engine.
Does eCommerce Insights write product content automatically?
No. Every recommendation ships as a reviewable diff — proposed title, description, bullet set, schema block, or metafield value shown next to the current one. A human on your team reviews, edits, and approves before anything changes. Bulk approval exists for similar diffs, but the review step never disappears.
Can it push changes to my store, or do I export?
Both. On Shopify, approved diffs push through the admin API on the Growth plan and up (Early access); every write is logged, attributed to the approver, and reversible. On every other channel, and on every plan, approved diffs export as CSV your team can ship through its own workflow.
Does it work if I'm not on Shopify?
Yes. eCommerce Insights is Shopify-first but channel-aware: Amazon catalogs get Rufus and COSMO scoring with A+ Content and backend-keyword recommendations, and Adobe Commerce, BigCommerce, WooCommerce, and headless storefronts get the six-engine scan with vendor-specific output. Recommendations always match the channel they ship to.
Reference: the diffs recommend against the schema.org Product vocabulary — the same structured data AI engines and shopping agents parse.
Ready when you are
See both scores on your own catalog.
14-day free trial. Six engines. No credit card.
2,500 products · 6 engines · daily refresh on Growth