All this while, the Sitecore community had been trying to bring in Sitecore ML in the former versions, but luckily, now we have Sitecore Cortex that offers a skeleton making it easier to incorporate machine learning models within content authoring and marketing workflows.

Marketers today are empowered with the ability to understand user behaviour and predict relevant content to the corresponding persona.

With the latest version of Sitecore in place, i.e. Sitecore 9.1, the machine learning algorithms, as well as Natural Language processing (NLPs), are used to study the patterns of user behaviour to develop ML models for prediction purpose.

Sitecore Version 9 is a “one connected platform” and has the ability to tag, track, measure and optimize the customer digital touchpoints or end-to-end journey in a lucid way. 

The situation till now

While Azure Machine Learning increases site engagement by helping companies to deliver the most pertinent content to the customer, most organizations struggle with maintaining multiple platforms serving different parts of their customer journey, making it all the more difficult to build an accurate picture of customer behaviour and opportunities for improvement.

Why Did We Choose Azure Machine Learning 

If you’re looking to drive even more business value from your xDB data, experimenting with ML models is now easier than ever. 

Sitecore Cloud on Azure offers excellent performance in deploying and scaling infrastructure on-demand while significantly reducing development costs. Using native Azure services and a pay-as-you-go usage-based model can simplify IT operations.

Let's now take a close look at the six reasons for using Azure Machine Learning and Sitecore for one-to-one personalization:

1. Built for the cloud-first mobile-first world - There are specific things that machines do better than humans. These include analysing data and finding patterns within large datasets to make close to accurate predictions.
2. Easy for your team to draw the right skillset - We have realized that it unnecessary for the data scientists to delve into the marketing context.

    This is how the skillset mix works:

• Data scientist with platform or R programming expertise and ability to calibrate models for best performance
• Sitecore analyst deeply understands xDB and xConnect data compositions
• Sitecore developer who understands the pipelines and solution architecture
• Business/marketing stakeholder understands marketing goals, audience demography and KPIs for the organization
3. Visual composition, easy to use, drag and drop – With the help of Azure ML, you can perform some actions on the basis of the algorithms without having to code but just by drag and drop the said functionality. This is possible because Azure gives us a large verity of in-built tools and packages.

                                         The drag and drop facility in the Microsoft Azure ML Studio

1. Capability of using different data sources – Now you can pull your data from various sources and formats such as csv file, Azure Blob, RDBMS, Hive, Web.
2. Content Tagging - Content tagging is freshly introduced in Sitecore 9.1. It's a feature that enables integration of Sitecore CMS with machine learning (ML) based natural language processing (NLP) engines like Open Calais.

                                                     Content Tagging Process in Sitecore Cortex

Read our blog: How Automation of Content Tagging in Sitecore Cortex works in the presence of Azure ML Studio.

Webinar On-Demand

1-1 Personalisation using Azure ML and Sitecore: A step towards better user experience

Register Now

1. Deploy in Minutes – One of the best features we found in Azure ML is it was done in the cloud [1] and it wouldn’t hamper your work once you’ve built the model. The model is itself built on Azure ML and it is provided to you as a web service. You can continue to work on your solution as well as Azure ML simultaneously. Thus, you can deploy on Azure Machine Learning Studio and get the results and send results thorough an API.

What the Future Holds

For now, we have just used the xDB data along with the tags data. In case we want to define the scope of Personalization using Azure Machine Learning in the future, the following features can be brought to use to mileage the one-to-one personalization with the viewer.

• Combining Multiple Features (eg. Tags, Geo Location)
• Refine Custom Rules – We can use tags and certain other things in Azure ML for better personalization as against use of just cluster numbers now
• Search Using ML and NLP – Making use of SQL ML to recommend content to the user based on his digital search behaviour. This will make content tagging completely automated.
• Data from all channels – You can fetch data from all the sources such as the web, social media, salesforce, etc. and create accurate customer profile.

Altudo is a Sitecore Platinum partner with 500+ Sitecore projects delivered for 45+ Fortune 500 brands, across 7 industry verticals
We help you realize the true potential of your marketing efforts, speeding up the Time to Value by leveraging CX strategy, 1:1 personalization & our global delivery expertise.

Ashish Kapoor Director Technology Solutions

• In
• Em