New for Marketers in Sitecore 9.1: Cortex - Machine Learning

In my earlier post What's New for Marketers in Sitecore 9, I summarized the new features in Sitecore 9 and 9.1 that are most relevant for marketers. In this post I'll go into more detail on Sitecore Cortex.

Released as part of Sitecore 9.1, Cortex is the unified name for the Sitecore features that combine to enable “Contextual Intelligence”:

1. xDB – data collection and processing.
2. xConnect – data to and from devices, campaigns, or third-party sources like CRMs, POS, or ERPs.
3. Machine learning – smart algorithms process your data, surface opportunities and trigger intelligent responses.

As you can see not all of it is new. xDB has been powering advanced features in Sitecore since version 7.5. xConnect was released with Sitecore 9. Therefore, marketers who have been using Sitecore xDB and xConnect will have large volumes of customer behavior data. The reality is that the amount of data collected can be overwhelming after a while, and is way too much to process manually.

AI Queries and Machine Learning algorithms take the manual work out of analyzing collected data, and personalizing user experiences based on that data. Machine Learning is first configured by a person and then run for a period of time. Periodically the data models need to be analyzed again, preferably by a data scientist, to determine if predictions have been accurate or not.

Let's go over the steps in more detail:

1. Collect the data – Start collecting meaningful data for the business with xDB as the repository. xConnect will help you feed data from third parties to xDB. But enabling xDB is not sufficient; to determine what data you want to collect you must use Sitecore's digital marketing features and configure:
   • Profiles
   • Profile Keys/Personas
   • Profile Cards
   • Patterns Cards
   • Goals
   • Events
   • Campaigns
2. Polish the data – Just collecting the data is not enough, you must have lots of relevant and recent data. You need to continuously review your data to make sure that you are not collecting data that is not relevant to your business. Use Sitecore's marketing tools to keep your data relevant:
   
   • Assess visitor engagement – Adjust visitor scores accordingly.
   • A/B and multivariate testing – Find out what content converts better.
   • Rules-based implicit personalization – Tailor content based on geography, demographics, behavior, etc.
   • Explicit personalization – Use the data in your CRM to target content.
   • Marketing automation / Engagement plans – Automate marketing post initial contact.
   • Path analyzer – Optimize pages on your conversion funnel.
   • Experience analytics – Continuously monitor your efforts.
3. Model and process the data – Once you have collected relevant data in gigs, it is time to automate the digital marketing operation. You'll need:
   • A Machine Learning service (from Microsoft or other vendors like IBM) – Developers will need to work with data scientists and Sitecore analysts to be able to develop and verify the model.

As you can see there is a bright future for Sitecore Cortex™ in marketing automation, but getting it to work requires time and effort. This can be the right choice for you if one of your business needs is strong personalization capabilities and you have thousands of website visitors and thousands of content options. Machine Learning can help you identify key audience segments and make relevant recommendations to your visitors. Some of the wonders Cortex can do for you include:

1. Assessing visitor engagement to identify visitors with the highest prospect of converting.
2. Offering personalization suggestions to target leads with relevant content.

In the CMS there is a new tab on the Experience Optimization tool, where Sitecore Cortex™ personalization suggestions can be found. When a component test is finished and the winner is selected, you can see suggestions about how to personalize the page.

Personalization Suggestions powered by Sitecore Cortex™

Sitecore's promises for this intelligent assistant of the future are:

1. Profiling customers based on content consumption.
2. Building smart distribution models.
3. Driving self-learning engagement scoring.
4. Powering automatic content tagging and intelligent recommendations.

For more on Sitecore 9's features for marketers, see my previous post: What's New for Marketers in Sitecore 9?