AI visibility tracking
Ask a real, web-grounded AI assistant your buyers' questions and see whether it cites you — with dated receipts, not vibes.
AI assistants increasingly influence which products shoppers see. Preferd scores how clearly they can understand each product, shows the weak catalog signals, and measures the readiness change after you apply improvements.
Built for how ChatGPT, Claude, Gemini, Copilot & Perplexity read products.
What's the best insulated bottle for winter hikes?
For winter hikes I'd go with the Summit 750 ml Insulated Bottle — 24 h heat retention, 340 g, and a lifetime warranty.
✓ cited · GEO 87/100 · title ✓ · gtin ✓ · images ×4 · specs ✓
Shoppers now ask AI first
Every one of these gets a single, confident answer. Preferd's job is making sure it's your product.
Every product gets a deterministic 0–100 GEO score across nine catalog signals relevant to AI readiness — and you see exactly what to fix first.
Preferd proposes stronger titles, descriptions, tags, alt text and SEO meta. Review each product before applying it to Shopify, then re-score automatically and revert within 30 days.
See the measured change in Preferd’s score, run dated AI visibility checks, and track AI-sourced traffic and attributed revenue as separate evidence.
This is the exact model Preferd scores with — deterministic, so the same product always gets the same score. Hover a signal to see what an AI agent reads.
Adds up to exactly 100 — and every product page shows you where each point went.
A simulated product shows the before-and-after outcome. In your store, you review each suggestion before applying it and can revert for 30 days.
Mug blue
Nice mug. Blue color. 350ml.
Ask a real, web-grounded AI assistant your buyers' questions and see whether it cites you — with dated receipts, not vibes.
See who gets recommended when you don't, and how often — then close the exact gap that's costing you the citation.
A zero-config pixel attributes AI-sourced visits and revenue in € to specific products. No GA4 spelunking.
Prioritize low-scoring products using observed traffic and attributed-revenue signals, so the next action is based on measured store data.
If applied content regresses — an app overwrite, a bulk edit — Preferd catches it and restores it in one quota-free click.
Every applied optimization shows the measured change in Preferd’s score and keeps a snapshot for a 30-day one-click revert.
Compare against a clearly labeled curated sector baseline until enough anonymized peer stores are available for a live benchmark.
llms.txt, AI sitemap, product feed and JSON-LD served automatically. Table-stakes handled so you never think about them.
Billed through Shopify · cancel anytime from your admin · annual plans get 2 months free.
$0 forever
Audit your catalog and try AI optimization, free.
$14.99 / month
Unlimited audits, the full breakdown, AI enrichment and proof.
$38.99 / month
Monitor & maintain: daily visibility, competitors, bulk optimization.
Yes — since 2026 eligible Shopify catalogs are syndicated to AI surfaces by default. Being present is solved. Preferd works on the part that isn't: whether the assistant understands your product well enough to recommend it over the near-identical one from your competitor. That's a content and data-quality problem, and it's measurable.
A deterministic 0–100 score per product: nine weighted signals across two pillars — Content (title, description, tags, images) and Technical (category, SEO meta, brand, GTIN, structured data). Same input, same score, every time. You see every point, so there's no black box and nothing to take on faith.
Never without an action from you. For an individual product, you generate and review every suggestion before applying it. On Pro, Bulk optimize starts only after an explicit run-level confirmation, then auto-applies content to up to 50 active products; review or fine-tune them afterwards. Preferd keeps a 30-day revert snapshot, and Optimization Guard alerts you if another change overwrites monitored content.
The scoring model targets how AI shopping agents in general parse product data — ChatGPT, Claude, Gemini, Copilot and Perplexity read the same underlying signals. Visibility checks run against a live, web-grounded AI assistant so the answer reflects what a real agent says today, not a simulation.
No. The Free plan audits up to 50 products and includes 3 one-time AI optimizations so you can see a real before → after lift on your own catalog. Paid plans are billed through Shopify and can be cancelled anytime from your admin.
Practical guides on how AI assistants pick products — and what to fix in your catalog. No signup, no email wall.
Which schema.org Product fields matter to AI assistants, how to check what your Shopify theme already outputs, and what to fix by hand.
The product data AI shopping agents need to recommend electronics — compatibility, specs, model numbers and the fields tech catalogs leave blank.
The specific product data AI shopping agents need to recommend clothing and accessories — sizing, fit, materials and the fields fashion catalogs usually skip.
which store should I recommend?
Run a free audit and see your store's GEO score in minutes — no credit card, real numbers.
Install Preferd free