Predict forecasts how likely users are to perform an event in your application.
Predict grades users based on their previous behavior. Users who are likely to perform an event receive an A, and users that are unlikely to perform an event receive a D. Predict scores are updated every 24 hours for active reports.
Additionally, Predict goes beyond the grade and indicates how likely it is your users will convert by calculating lift. For each letter grade, the lift at any given time indicates how much more (or less) likely a customer with that grade will convert vs. the average customer (ie. an “A” customer with a 40% chance of conversion will have a lift of 4 if the average customer has a 10% chance of conversion).
When a user completes the goal event, they are counted in the number of converted users.
Building a Predict Model
To create a predict job, you must first determine your conversion event goal. To create a new predict job:
- Go to Predict.
- Click Create New if you have existing Predict jobs. You are directed to the Predict builder if you have never used Predict before.
- Name your Predict job.
- Select a goal event.
- Select if this user is doing this for the first time ever or not. By default, the "for the first time ever" box will be checked. If selected, Predict will analyze users who have never performed the event. If not checked, Predict will analyze all users, regardless of whether or not the event has been performed before.
- Click Save prediction. If a pop up window prompts again, click Save prediction.
- A notification email is sent once the Predict job has finished running.
Manage or Delete Existing Predict Jobs
It is possible to run up to ten different Predict jobs at a time. To manage and delete Predict jobs:
- Go to Predict.
- Click the All Reports icon. Click All to view all Predict jobs for the project, or click Your Predictions to see Predict Jobs that you have started.
- To delete a Predict Job, hover over the job and click the trash icon that results.
Predict Model Degradation
While rare, it is possible for successful predict models to degrade to the point of no longer being predictive. This will happen if customer behavior drastically deviates from the behavior used in the existing model or if data quality degrades.
If this happens, Mixpanel will automatically turn off the model, send you an email notification alerting you of the change, and turn off any existing notification campaign that uses the Predict grades.