Introduction to Predict

[#1]

This article describes the Tealium Predict ML product and how it is used to create, train, and deploy machine learning models to make predictions about visitor behavior.

In this article:

Table of Contents Placeholder

How it Works

Tealium Predict is a built-in machine learning technology product used with the Tealium AudienceStream CDP product that takes a complicated discipline (machine learning) and distills it down to a fully-managed feature. Coupled with AudienceStream, the product allows you to build and create more intelligent audiences and activate trustworthy machine learning insights across your entire tech stack.

Using Tealium Predict, you can create, optimize, and deploy data models that make meaningful predictions about your visitor behavior. Deployed models are managed by Tealium in real-time using a predictive output attribute that is made available to you after your model is trained and deployed. In just a few clicks, you can then create a model using the new output attribute, which is immediately available and actionable in AudienceStream.

Key Benefits

Key benefits of Tealium Predict include the following:

• Leverage Machine Learning Intelligence Across your Entire Implementation
  You can automatically activate predictive insights in any of Tealium's over 1,200 integrations.
• Simplification
  Tealium Predict simplifies the implementation of machine learning for you without the need for a Data Scientist or dedicated team. You can create a model, deploy it, and immediately activate the value in AudienceStream without writing a single line of code.
• Predict Customer Actions
  Automatically draw conclusions about what actions customers are likely to take in the future and design tailored programs that directly address their needs.
• Personalization
  The Tealium Customer Data Hub allows you to pinpoint meaningful visitor personalization data.
• Reduce Time Spent on Machine Learning Projects that Involve User Data
  Data collection and data activation are natively integrated with machine learning intelligence, which greatly reduces time spent for data preparation and activation.
• Score Customers with Predictive Potential
  Technology built into the model produces a score for visitors with the likelihood to return or complete a specific behavior. You can then leverage the scores to generate audiences for further proactive action, such as re-engagement campaigns.

Choosing What to Predict

A target attribute is a AudienceStream attribute that defines or signals the visitor behavior that you seek to predict with any Tealium Predict model. The target attribute must be either a Boolean/Flag or a Badge type attribute and be Visit or Visitor-scoped.

Selecting a Target Attribute

When you create a model using Tealium Predict, potential target attributes from your AudienceStream dataset display. To the right of each attribute, a rating of Ready or Not Ready displays to signify whether or not the attribute selected is ready for training. This rating simplifies model creation by clarifying which of your flags and badges are ready to be used to create models. A rating of Not Ready does not mean that the attribute is problematic in other contexts. It simply means that this attribute is currently deemed insufficient for successful training of a Tealium Predict model.

Machine learning technology in general requires a relatively high volume of data to succeed and machine learning models provide better results when trained on a large amount of data.

The following two factors to define target attributes that are Ready or Not Ready:

• The volume of data for the attribute
• The distribution of that volume between true and false values.

Both the true and false groups must be above a minimum threshold. For example, for the dates that span the Training Date Range, the median daily counts of True and False visitors must be greater than or equal to 200. This threshold is intentionally set as low as possible to provide the most options possible for target attributes. A model that uses a target attribute that is labeled Not Ready typically fails during the training process due to insufficient data for the model to learn on.

If the target attribute you want to use is labeled as Not Ready, try one of the following common solutions:

• Wait for more data to accumulate. The problem often solves itself over a longer time period.
• Identify ways to drive additional traffic to the AudienceStream data source.
• Add additional data sources to your AudienceStream profile so that the volume of daily data increases for this target.

Learn More About the Technology

To learn more about machine learning technology, see Machine Learning Concepts and Technology.