What is Repeat Buy and how does it work?

The Relo Repeat Buy predictions are designed to prompt one-time customers to order again by using precisely timed insights. Predictive reordering automates the repurchase reminder processes by triggering an event in Klaviyo making it easy for the merchant to use actionable insights in existing flows. Customers are targeted at the most impactful time based on personalised average reorder gaps helping increase conversion rates.

  1. We predict exactly when customers are likely to buy again.
  2. You can integrate this data with Klaviyo in seconds.
  3. Once activated, data syncs automatically with your existing email / sms flows.
  4. Use this actionable data to target customers better and personalise comms.
  5. Convert more customers in 2-clicks using Magic Cart. 

After activating the integration, the trigger "Reorder prediction by Blueprint" will show in your Klaviyo trigger list. 

Simply select the prediction and place at the start of a flow. This will usually replace an existing 'order placed' trigger for example.

How does the prediction work

Using the Relo > Shopify integration, we are able to access the purchase history of all your customers and make some calculations to create an individualised reorder predictions. We calculate:

  • The average order gap of the customer purchasing
  • The average order gap of the product purchased
  • The average order gap between orders across the merchant

If there has only been 1 x previous purchase from the customer, the reorder prediction will be generated on the average order gap of the product purchased. 

Messages are therefore sent to customers at the most optimal reorder times. Eg. when they’re most likely to repurchase.

We automatically group re-order prompts together. This saves on comms numbers and maximises efficiency.

Why use Relo predictions?

  • Relo combines merchant, product and customer level data to provide the most accurate predictions.
    • EG: Customer A buys Product 1 and Product 2 at different rates and different times. The reorder model will predict when and what they will buy next based on the merchants’ average customer reorder gap, Product 1 and 2s average reorder gap, and Customer A’s average reorder gap. This is done on a weighted, moving averages model that softens anomalies.
    • The more historic data a customer has, the more powerful the Relo predictions are. For new customers, the model will default to merchant and product level data.
  • Relo prioritises predictions based on expected ROI of each message type (reorder opportunity or upgrade to subscription) at any given time to make sure customers are only targeted with the best comms and never “spammed”.