Predictions

Predictions is a codeless AI-powered tool built on top of Synerise analytics that predicts any type of event or action on a customer’s journey.

Key features:

  • Predictions built on data already stored inside the Synerise platform
  • No coding required
  • No need of manual feature selection or creation
  • Fully customizable from the user’s perspective
  • Predictions are saved as events which ready to use further on (for example in the Automation module, the Communication module, and so on)
  • Easy way to discover top contributing factors to the entire model but also to single predictions

How long does the system calculates a prediction?
It depends on the system overload, but to give you an example how long you may wait, a prediction for a 2 mln customer group takes about 2 hours.

Requirements


  • To make use of this feature contact the Synerise Support in order to switch it on for your business profile
  • Create an expression that will be used to define what kind of information you want to get from prediction

Configuring a prediction


To configure prediction, you have to perform the following steps:

  1. Go to Image presents the Prediction icon > New analysis.
  2. Define the type of the prediction
  3. Select customers for whom the prediction is to be performed
  4. Select what kind of information you want to get from prediction
  5. Select additional data for the prediction model
  6. Configure the final settings

Define the type of the prediction

Prediction types
Prediction types

You can select one of two prediction types:

  • Customer regression is suited for perfoming the analyses that return the numerical results. It’s best used in cases such as:
    • Predicting the amount of money spent by particular group of customers in the defined time range
    • Predicting the amount of items purchased in the defined time range
  • Custom classification is suited for performing the analyses that return the true/false (or 1/0) values. It’s best used when you want to get to know:
    • Will a customer belong to a particular group of customers?
    • Will a customer leave in the next 30 days?

Select customers


Selecting an audience for prediction
Selecting a group of customers to make prediction

Select the audience for which you want to prepare a prediction.

  1. In the Audience section, click Define.
    • To select an existing group of customers, select the Segment tab (default) and select the groups. If you select more than one, the dependency between them is described by OR, which means that to receive an email, a customer can be only in one of the selected segments.
    • To define a new group of customers, select the New audience tab.
    1. In the New segment field, enter the name of the group of customers.
    2. Follow the procedure described here.
  2. Confirm by clicking Apply.

Select area of prediction


Selecting an expression to define the scope of prediction
Selecting an expression to define the scope of prediction

Select the expression based on which the prediction will be made. For example, if you select an expression that predicts Email OR in last 30 days, you will get prediction for OR for the next 30 days.

  1. From the dropdown list, select the expression you previously prepared.
  2. Confirm by clicking Apply.

Select model inputs


Selection of features to support predicting model
Selection of features to support predicting model

Select the features that the system will use to support the prediction process. A feature is a variable or an event that can be used by AI engine to make a prediction. The list already contains predefined features.

  1. Click Add feature.
  2. From the dropdown list, select one of two options:
    • Manually - You can pick features on your own.
    1. On the list, select the checkboxes next to the features you want to include.
    2. Confirm your selection by clicking Add.
    • Automatically - All features are added to the list (recommended).
  3. After making a selection, click Apply.

Feature analysis

Selection of features to support predicting model
Analysis preview in the form of a line chart

The unit of the feature depends on the type of the feature. For the features that are events, the unit is an occurrence of an event.

Assuming that a feature is an event, then:

  • Count - The number of the event occurrence
  • Min - The minimum number of the event occurrences
  • Max - The maximum number of the event occurrences
  • Missing - The number of customers for whom the feature is not counted
  • Mean - The mean occurrence of the event

Settings


Selection of features to support predicting model
Analysis preview in the form of a line chart

In this section, define the frequency of recalculating the prediction and settings of the event that is generated for a customer for whom the prediction was made.

  1. In the Model configuration section, select the number of days after which the prediction is recalculated.

  2. From the How far in advance do you want to make a prediction? dropdown, select the number of days for which you want to make a prediction (calculated from the current date).

  3. To switch on the repeatable calculation of prediction, select the Enable repeatable calculation checkbox.

  4. In the How would you like to display results? section, select the scale of prediction results - two- or five-point scale. The scale is represented as a score_label parameter of the snr.prediction.score event.

  5. Use the slider to define the number of features displayed in the parameters of the event generated for a customer for whom the prediction was made.

    Note: One of the features is a prediction parameter, it is included in the snr.prediction.score event and it can be used while selecting customers for your campaigns.

  6. In the Name field, enter the value of the scoreName parameter of the event generated (snr.prediction.score) when a prediction is made.

    WARNING:

    You can use the following characters:

    - a-z - only lower case
    - 0-9
    - special characters: `.` and `_`
    

    generic scoring event
    The preview of the prediction generation event
  7. Click Apply.

  8. To save:

    • as a draft, click Save.
    • and calculate, click Save & Calculate.
      Result: The output of the prediction is an event generated on the customer’s profile.

Actions on the prediction list


In the Predictions module, on the predictions list, you can perform the following actions for one prediction at a time:

Action name Description
Calculate Immediately starts the first calculation of the prediction.
Recalculate Immediately starts the subsequent calculation of the prediction.
Edit Opens the editing mode of the prediction.
Duplicate Creates a draft copy of the prediction.
Delete Deletes the prediction from the list, however, the events generated by the prediction are kept.
Stop Stops the calculation of the prediction

Prediction statuses


The predictions can have the following statuses:

Status name Description
Calculating The prediction is not ready, the calculation is in progress.
Recalculating The prediction is being recalculated due to retraining model settings enabled in the configuration of the prediction.
Done The prediction is ready.
Draft The prediction is saved as a draft and not ready to be submitted for calculation.
Issue The prediction passed the training stage, but only returned one answer for every row in validation set.
Error The prediction didn’t pass the training stage, try to submit it for calculation once again or contact the Synerise support team.
Outdated The prediction cannot be calculated due to an error - in such case contact the Synerise support team.
😕

We are sorry to hear that

Thank you for helping improve out documentation. If you need help or have any questions, please consider contacting support.

😉

Awesome!

Thank you for helping improve out documentation. If you need help or have any questions, please consider contacting support.