The insights derived from the "Predicted Time Until Next Order" report allows the user to more effectively channel resources towards accelerating purchase timelines by targeting different "days until order" and "order likelihood" customer cohorts.
The report's data
The report's output is a combination of two different modeled attributes — "Order Likelihood" and "Days Until Next Order." The models look at the last 180 days to predict the next 42 days (6 weeks). Each model is custom built to the data source, retrained monthly, and new predictions run nightly.
Review this guide to learn more about the data science and modeling behind the report.
Explore the data
Refine and explore the order likelihood report in four ways:
- Change the predicted time until order between days and weeks.
- Select or deselect likelihood thresholds.
- Select or deselect chart sections.
- Select or deselect the period.
The additional insights and sample of customers below the chart will update to match your selections.
Leverage the report's data in two locations:
Directly within the report
Create a segment matching the displayed criteria after manipulating the data via the report explorer. Simply click the Save As Segment button to name and create the segment. Once saved, you'll be redirected to the campaign builder to start using the segment immediately.
In the segment builder
Information from the report is available in the segment builder, using the Order Likelihood and Days Until Next Order attributes.
Example use cases
Typical use cases for this functionality include:
Use "Order Likelihood" to help accelerate revenue among the "likely to buy" customer cohort.
- Consider grouping with other always-on revenue acceleration campaigns like Browse Abandonment or Cart Abandonment in a priority group and delivering a campaign to Very Likely or Likely to buy customers who are not actively shopping the site.
Use "Order Likelihood" to find new customers who look like existing ready-to-buy customers.
- Sync a segment of Extremely Likely to buy customers to Google or Facebook for lookalike modeling, to find new customers who look like them but for whom your brand is not in their consideration set.