AnswerRank

Help Center

Common questions about getting started, scoring, plans, and methodology. Can't find what you need? Email support@answerrank.com and we'll reply within 24h.

Getting started

  1. Install AnswerRank from the Shopify App Store.
  2. We sync your catalog automatically and run the first audit (60–120s).
  3. Open the Overview dashboard to see your AI Visibility Score, Quick Wins, and the rank tracker.
  4. Add 5–10 buying-intent prompts on the Prompts page (or use the suggested starter pack for your category).
  5. The first rank-check fires within 24 hours. Click "Run scan" to trigger it sooner.

Adding your first prompts

Good prompts are questions or queries your customers would actually type into ChatGPT, not keyword soup.

✅ Good: "best organic cotton t-shirts under $50", "affordable merino wool sweaters made in USA"
❌ Bad: "t-shirt", "buy clothes"

You can bulk-import via CSV with columns prompt, category, volume_label. Volume label is your own estimate ("low", "med", "high") — we don't query external keyword tools.

Understanding the audit score

Every product gets a 0–100 score from four dimensions:

  • Schema completeness (30%) — product type, vendor, image alt text, SKU, GTIN/barcode, weight, ≥3 tags, ≥1 metafield, prices, compare-at-price.
  • Attribute coverage (30%) — % of category-specific attributes present (apparel: material, fit, country of origin, certifications…).
  • Description quality (25%) — 150–500 words, spec list <ul> or <table>, FAQ block, front-loaded attributes, ≥1 quantitative spec.
  • Image quality (15%) — ≥3 images, all alt-texted, at least one alt text contains a factual attribute.

Score = 0.30 × schema + 0.30 × attribute + 0.25 × description + 0.15 × image.

Reading the rank tracker

The Overview → Rank tracker table shows your latest position per (prompt × engine):

  • Green badge #1–#3 — top result. AI engines are likely to cite you first.
  • Yellow badge #4–#10 — in the answer but not top.
  • Grey — — not mentioned in this engine's answer.

Position 1 doesn't mean a deterministic AI rank — it means you were the first brand mentioned in the engine's natural-language answer. Brand order in an AI answer is the closest signal to "rank" that exists today.

Plan limits and upgrades

PlanPricePromptsEnginesDaily API cap
Free$05ChatGPT + Gemini
Starter$39/mo25ChatGPT + Gemini$0.50/day
Growth$99/mo100ChatGPT + Gemini + Competitors + Bulk rewrite$2.00/day
Pro$249/mo500All Growth features + A/B testing + priority support$10.00/day

All paid plans have a 14-day free trial. Upgrades take effect immediately (Shopify prorates). Downgrades take effect at the end of the current cycle. Cancellation downgrades you to Free; no data is deleted.

GDPR / data deletion

We store no customer PII. Order data is stored as aggregates (total, currency, item count, source) — no names, emails, or addresses.

To delete all your data: uninstall AnswerRank from Shopify Admin → Apps. Shopify fires shop/redact 48 hours later, which permanently deletes every row associated with your shop.

To export your data first: Settings → Data & Privacy → Download JSON.

Troubleshooting common issues

"My dashboard shows zero rank data" — Click Run scan on the Overview page. Adding a prompt creates the row but doesn't auto-scan. If you've already scanned and still see no data, the day's scan limit may be reached — new scans resume tomorrow at 00:00 UTC.

"Scan finished but I'm not in any answer" — That's actionable data, not a bug. Check the Competitors page to see who is being cited and what attributes they're cited for. The Quick Wins panel tells you what to fix.

"Quick Wins look stale" — Click Settings → Data & Privacy → Re-run audit. Quick Wins regenerate from the latest AuditScore data.

How we track each AI engine (methodology)

ChatGPT: OpenAI's Responses API with the web_search_preview tool. We submit your prompt verbatim and parse the model's natural-language answer for brand mentions.

Gemini: Google AI Studio's gemini-2.5-flash with googleSearch grounding. Same prompt, same parse logic.

For both engines, we extract brand citations using Anthropic's Claude Haiku (with regex-based fallbacks). Your prompt is the only thing sent to OpenAI/Google. Your catalog data is never sent to either.

Why this isn't "true rank": LLMs don't have a deterministic ranking function. The same prompt asked twice can produce different answers. We treat the first-mentioned brand as position 1, second-mentioned as 2, etc. — the best heuristic available.

Connecting Slack

  1. In Slack, create an Incoming Webhook for the channel where you want alerts.
  2. Copy the webhook URL (https://hooks.slack.com/services/...).
  3. Paste into Settings → Notifications → Slack webhook URL and click Save.

Alerts will fire on: rank drops (top-3 lost), new competitors entering top-5, AI-attributed revenue drops >20% WoW.

Exporting your data

Settings → Data & Privacy → Download JSON produces a one-file dump of everything we store for your shop: shop record, subscription, prompts, rank results, products, audit scores, competitors, quick wins, orders + line items.

For order-level CSV, use Attribution → Export CSV in the Attribution page. Same data, flatter shape, opens in Excel.