SoonRizecart reads your Shopify or Salla store the way ChatGPT, Claude, Gemini and Microsoft Copilot do, fixes what they can't understand, and tracks when they recommend you.
We check the answers on
30 seconds
Unedited screen recording. A buyer asks a full question and gets a short list of stores. There is no second page to be on.
Buyers describe what they want, in their own words, in Arabic or English.
The assistant recommends a handful of stores and products. Everything else is invisible.
The model can only recommend what it can read, match and trust.
The shift
Not a prediction. Published measurements from the last two years of retail traffic.
Measured by Adobe Digital Insights and published in Similarweb's 2026 generative AI landscape analysis.Read the source
The problem
There is no rank to check and no report to open. The answer is written, the buyer acts on it, and you never see the question.
What a merchant can see today
No data
AI referrals arrive as direct traffic. No keyword, no position, no impressions. The one channel growing fastest is the one you have no dashboard for.
The assistant recommends three or four stores. Being fifth is the same as not existing, and nobody asks for the next page.
Thin descriptions, missing barcodes, no structured data, JavaScript-only pages. The model cannot match a product it cannot parse.
Most stores publish one language properly. Ask the same question in Arabic and the shortlist changes completely.
Why SEO isn't enough
Search engines rank pages. Assistants read facts, then write one recommendation. The work is related, but it is not the same work.
Product feeds keep Google Shopping happy. They do not tell Claude who a product is for, or why it beats the alternative.
How it works
Three steps. Nothing changes in your store until you approve it.
Paste your URL for the free audit, or install the Shopify or Salla app for the full catalogue. We read, we don't write.
About a minute · read-only
A score out of 100, seven categories, and every issue in plain language: what's wrong, why it matters, how to fix it.
Plain language · technical details one click away
Fixes appear next to the current value. Edit, approve, publish to your store. Roll back any batch with one click.
Before/after · roll back anytime
The platform
Three screens do the whole job: find what AI can't read, fix it with your approval, then watch the answers change.
We crawl your store the way an AI crawler does and score seven categories out of 100. Every issue is written in plain business language.
example-store.com
Detected platform: Shopify
AI can tell this is a product page
Why it matters: Structured data is how you tell machines 'this is a product, here is its name, price and photo'. Without it, AI has to guess from page text — and often guesses wrong.
Recommended fix: Add Product structured data (JSON-LD) to every product page. Shopify themes usually include it; on Salla verify your theme emits it.
Your branded products carry a barcode AI can match
Why it matters: A GTIN/barcode lets AI connect your listing to the manufacturer's data, reviews and comparisons. Without it, a genuine Versace or Samsung product is treated as an unknown item.
Recommended fix: Add the GTIN (EAN/UPC) and manufacturer part number to every branded product. Rizecart can look them up and fill them in for you.
Product descriptions explain who the product is for and why to choose it
Why it matters: AI recommends products it can explain. A four-word description gives it nothing to say, so it picks a competitor with a fuller listing.
Recommended fix: Write 80–200 words per product covering what it is, who it suits, key specifications and what makes it different. Rizecart drafts these in your voice.
Example layout with sample values; the check titles and advice are the product's real copy. Run the free audit to see your real numbers.
Each one runs the way AI crawlers read your store, and each comes with a plain-language fix.
ChatGPT, Claude, Gemini and Microsoft Copilot each use their own crawler. If your robots.txt blocks them, your products cannot appear in their answers no matter how good they are.
Found by the audit, with a recommended fix
Structured data is how you tell machines 'this is a product, here is its name, price and photo'. Without it, AI has to guess from page text — and often guesses wrong.
Found by the audit, with a recommended fix
A GTIN/barcode lets AI connect your listing to the manufacturer's data, reviews and comparisons. Without it, a genuine Versace or Samsung product is treated as an unknown item.
Fixed from the Fix Center
AI recommends products it can explain. A four-word description gives it nothing to say, so it picks a competitor with a fuller listing.
Fixed from the Fix Center
Gulf shoppers ask AI in both languages — often English for brands and Arabic for everyday goods. A product that exists in only one language is invisible to half the questions.
Fixed from the Fix Center
llms.txt is an emerging convention that gives AI a concise guide to your store: what you sell, key categories, policies. Early adopters get read more accurately.
Fixed from the Fix Center
An honest look at the three ways merchants deal with AI search today.
| What you get | rizecart | SEO agency | Doing nothing |
|---|---|---|---|
| Checks your store the way AI crawlers read it | yes | partly | no |
| Findings in plain language, in Arabic and English | yes | Reports, usually in English | no |
| Fixes published to your store after your approval | yes | You or your developer apply them | no |
| Records what ChatGPT, Claude, Gemini and Microsoft Copilot actually answer | yes | Rarely | no |
| Time to first result | About a minute for the audit | Weeks to months | Never |
| Cost | Free audit, then plans by catalogue size | Monthly retainer | Free, until a competitor is the answer |
If you already have an agency you like, keep it. Rizecart gives them the audit and the monitoring; you keep the relationship.
Who it's for
You don't need to know what structured data is. You need to know which products AI can't read, and to approve the fix.
You added the products yourself and there is no marketing team. You want the list of what's broken and a button that fixes it.
Best on Starter
You have a catalogue, competitors you can name, and someone who owns growth. You want daily monitoring and the reason a competitor was recommended.
Best on Growth
You manage more than one brand, need API access and exports, and report results to someone else.
Best on Scale
When it isn't a fit
If you sell only through marketplaces and have no store of your own, we can audit the storefront but cannot publish fixes to it. Zid support is in development.
Connect in a couple of clicks. We read your catalogue, propose fixes, and write back only what you approve.
SoonZid support is in development. Tell us you need it and we'll let you know the day it's ready.
Ask about ZidPricing
Priced by catalogue size, published in full. Annual billing saves 20%.
Starter
$39/month
billed annually · Save 20%
1,000 credits / month
500 products
Growth
$79/month
billed annually · Save 20%
3,000 credits / month
2,500 products
Scale
$159/month
billed annually · Save 20%
8,000 credits / month
Unlimited products
Short answers. If yours isn't here, write to us.
No. The free audit reads your store from the outside, exactly like an AI crawler would. Connecting your store is only needed later, when you want fixes published for you.
Never. Every fix is shown as before/after, and nothing is published until you approve it. Every published batch can be rolled back.
Yes. that's the point. Rizecart is built Arabic-first for Salla and Shopify merchants in Saudi Arabia and the Gulf. Audits, fixes and monitoring work in Arabic and English.
We check crawler access and structured data for ChatGPT, Claude, Gemini and Microsoft Copilot, and run monitoring questions on ChatGPT, Gemini, Copilot and Claude depending on your plan. Copilot answers are measured through Azure OpenAI (the models behind Copilot) with web grounding — Microsoft offers no public Copilot API.
The score is measured from what we actually found on your store. Where an AI model's judgement is involved, we label it as AI analysis so you always know which is which.
Usually under a minute. Large stores with many products can take a little longer while we sample product pages.
Free, no signup, about a minute. Your report is yours to share.
Or see the three plans first