Use case

Abandoned cart recovery
one decision per shopper, not one schedule

Most filled carts never reach checkout. Purands decides for each shopper whether a reminder is worth sending, on which channel, and when.

The problem

A shopper fills a cart, gets interrupted, and never comes back. The order is lost to a moment, not to a decision about your product.

The usual answer is a fixed sequence: an hour, a day, a discount. It treats a browsing student and a repeat buyer alike.

What triggers it, and what it reads

The trigger is a started checkout with no order behind it. What happens next is decided from the customer profile, not from a queue position.

  • Purchase history and average order value
  • Browsing behaviour before the cart was filled
  • Channel affinity: where this customer actually replies
  • Purchase intent and churn risk, scored daily
  • Every earlier conversation, so no question is asked twice

Brands running the abandoned cart skill report recovery moving from an 8% baseline to 23%.

What gets sent

Examples only, written for this page. Real messages are generated per customer from their own profile and your catalogue.

  • WhatsApp: Hi Priya, your Glow Essence Toner is still in your cart. Want me to hold it while you decide?
  • Email: Opens on the item left behind, with the reviews that answer whatever the shopper stalled on.
  • SMS: Your cart is saved for 24 hours. Tap to finish: [link]
  • Push: Opens straight to the cart, not the homepage.
  • Web assistant: If the shopper returns, the assistant picks the thread up where the conversation stopped.

How the agent picks the moment and the channel

The agent learns each customer's peak engagement window and sends inside it, rather than at a fixed offset from the abandonment.

Channel follows the same logic. A customer who reads WhatsApp and ignores email gets WhatsApp, and the email is not sent as well.

Fatigue is judged across every channel at once. If this customer has already heard from you enough this week, the agent sends nothing.

A discount is not automatic. It goes to customers predicted to need one, which keeps the margin on the ones who would have bought anyway.

Setting it up

Connect the store and orders, customers and catalogue sync in. Shopify, WooCommerce, Magento, BigCommerce, PrestaShop, Wix, Squarespace and WordPress are supported.

WhatsApp needs an opted-in audience. Double opt-in flows are built in, template approval with Meta is handled by Purands, and existing WhatsApp Business numbers can move across.

The other retention jobs

All six run on one engine: Agentic Skills, reading Profiling.

FAQ

Abandoned cart recovery

That depends on the customer. Email has a one-hour recovery window; on other channels the agent sends inside the customer's own peak engagement window rather than at a fixed offset.
No. The agent decides per customer whether to send anything, weighing everything else that customer has already received across all channels that week.
No. Discounts and loyalty points are issued only to customers predicted to convert with one, so margin is not spent on shoppers who would have returned anyway.
Yes. WhatsApp and the web assistant are two-way: the AI agent answers product questions, applies codes and processes re-orders, escalating to a person only when it needs to.
See it on your data

See this running on your own data

A 30 minute demo on your catalogue and your customers, with the decisions the agent would make shown one by one.