Show every shopper their own products — on the site, in search and in emails
The same storefront for everyone loses sales: a loyal customer sees what a random visitor sees, and search fails on “dres”. The platform builds recommendations and search from each customer’s behavior.
Personalization: recommendations and search
- “Similar products” blocks are filled by hand or not at all
- Site search does not forgive typos or a wrong keyboard layout — the visitor sees “nothing found”
- Emails are the same for everyone, and nobody has measured the effect of recommendations
- 01
Upload the catalog
A product feed (YML, Google Merchant, CSV or JSON) on a schedule or via the API. Popular, viewed together, bought together, similar and personal models are recomputed daily.
- 02
Add widgets and search
A recommendation block is set up in an editor with a preview — design, display rules and A/B variants — and inserted with one line. The search box with suggestions handles typos, word forms and a wrong layout and ranks products by the visitor’s interests.
- 03
Put products into emails
Personal products are inserted at send time — in abandoned carts, after a purchase and in regular campaigns.
- 04
Collect reviews
After a purchase, a review request with a link. AI-checked reviews appear on the product page, and stars show up in recommendation widgets and emails.
- 05
Measure the contribution
Attribution shows how much revenue came after clicks on recommendations, search and campaigns — with last-click, first-click and linear models.
- Every visitor sees picks for their interests without a merchandiser’s manual work
- Search finds the product even with a typo in the query
- The contribution of recommendations and search to revenue is in one report
Built on
- Abandoned carts and repeat purchasesThe first automated scenarios for a store: carts, welcome, the purchase after the purchase.
- Win back customers who are leavingChurn prediction, RFM segments and reactivation journeys before the customer is gone.
- Loyalty programPoints, paying with points, tiers, refer-a-friend and promo codes — on the same profiles.
- Order notifications and cascadesTransactional messages with one API call and a WhatsApp → SMS → email cascade.
- Moving from another platformCustomers, events, consents and segments from Mindbox, Customer.io and Retail Rocket.
- Revenue and marketing attributionRevenue, LTV, traffic sources and sales attribution by channel — on the same events.
- Marketing with a small team and AIClaude or ChatGPT answer from reports and build segments, emails and journeys over MCP.
- Large databases and strict requirementsFrom 250,000 profiles: a dedicated IP, an SLA by contract, long retention and access control.
- Agencies and integratorsThe platform under your brand: your domain, client projects and your journey templates.
Try it on your own data
14 days on the Growth plan, no card required: up to 1,000 profiles and 500 emails a day.