Skip to main content

Replenishment

Replenishment is Unveild's family of predictive AI flows for previous purchasers. Four metrics drive ESP flows under this umbrella, each fired at the moment Unveild estimates the customer is most likely to take that specific action:

MetricTriggers (in Klaviyo)What it's for
Reorder Recommendation UnveildReplenishment flowPrompt purchasers to reorder at their predicted reorder window
Push To Subscribe Recommendation UnveildReplenishment flowPrompt purchasers to convert from one-time orders to a subscription
Cross-sell Recommendation UnveildCross-sell flowRecommend complementary products to past purchasers
Winback Recommendation UnveildWin-back flowRe-engage lapsed customers to encourage renewed intent

Each metric carries the products Unveild recommends for that subscriber. The metadata structure mirrors Unveild's standard abandonment metrics, so existing flow templates can be reused.

Predictive Reorder​

Reorder Recommendation Unveild identifies the optimal moment to prompt each previous purchaser to buy a product again, on a per-customer, per-product basis. The score combines three inputs:

  • Intent Analysis — on-site behavior in the current session used to detect purchase intent.
  • Individual purchase history — each customer's past order patterns, used to determine average reorder intervals on a product-by-product basis.
  • Product re-order trends — average reorder timelines on a product-by-product basis across all customers who purchased the same product.

The metric fires at the moment the customer is most likely to repurchase — not at a fixed interval after the last order.

Push-to-Subscribe​

Push To Subscribe Recommendation Unveild identifies the optimal moment to prompt each previous purchaser to convert from one-time orders to a subscription. It shares the scoring family used by Predictive Reorder, with one additional subscription-specific input:

  • Intent Analysis — on-site behavior in the current session used to detect purchase intent.
  • Individual purchase history — each customer's past order patterns, used to determine both the average number of purchases before subscribing and the typical reorder intervals on a product-by-product basis.
  • Product re-order trends — average reorder and subscription timelines on a product-by-product basis across all customers who purchased the same product.

For each eligible purchaser, the model routes them to one metric or the other: customers scored as likely to subscribe receive Push To Subscribe Recommendation Unveild; customers scored as likely to reorder but not yet to convert to a subscription receive Reorder Recommendation Unveild.

Cross-Sell​

Cross-sell Recommendation Unveild increases customer lifetime value by determining the optimal time to cross-sell to previous purchasers, and which specific products to recommend. It uses three modeling approaches:

  • Affinity Modeling — identifies each customer's unique product preferences to recommend items they're most likely to purchase.
  • Lookalike Modeling — recognizes patterns among similar customers (e.g., people who buy Product A often go on to buy Product B), then recommends Product B.
  • LTV Analysis — prioritizes cross-sells that historically lead to higher lifetime value.

Win-back​

Winback Recommendation Unveild re-engages lapsed customers — purchasers who have stopped buying for a meaningful stretch but show renewed intent (e.g., returning to the site, viewing products) — at the moment they're most likely to convert again rather than on a fixed days-since-purchase schedule.

Setting up the flows​

Don't add wait times in any of these flows

Do not add any wait time before the first message in a Replenishment, Cross-sell, or Win-back flow. The metric is triggered for each individual at precisely the moment they're most likely to act — adding wait time pushes the message past the predicted window and degrades performance.

Requires:

  • The base Unveild tag installed
  • A configured ESP integration (currently Klaviyo)
  • A Klaviyo flow for each metric you want to enable, triggered by the corresponding Unveild custom metric

Per-category scoring rules, eligibility windows, and metric configuration are handled by your Unveild POC.

Who it's for​

Brands selling consumables, subscriptions, or repeat-purchase categories — beauty, supplements, pet, household goods, food and beverage — where reorder timing, subscription conversion, and lapsed-customer recovery drive long-term LTV.

Questions?

For setup help, reach out to your Unveild Customer Success Manager or saumo@unveild.ai.