Einstein Analytics Cloud Review

In this blog I will review the current status of two very interesting products of Salesforce Analytics Cloud: Einstein Analytics, which is about data exploration and dashboarding, and Einstein Discovery, which provides AI-powered insights and recommendations about your data to help understanding relations and trends and to predict what is going to happen next.

Einstein Analytics

Einstein Analytics dashboard example 

A lot of positive things are happening within Einstein Analytics. With many new and improved features in the Summer ’17 and Winter ‘18 releases I think this product has become very powerful. The top arguments to me:

  • The platform basics are taken care of: performance is great thanks to the concepts of replication and dataflow, and data security has been improved as well. Next to the security predicates, which are a bit complex to implement, there is now a simple checkbox to enable the Salesforce security model in Analytics. This provides additional options to only show the data a user needs to have access to. It makes a lot of sense to me to integrate the extensive security model of the Salesforce platform in Analytics Cloud.
  • If you were familiar with SAQL and JSON, many options in terms of visualization and data binding were already available. However, in each release parts of those programmatic tools have become available through the user interface. This means that more cool things can be done without any programming needed. Since there will now probably be multiple solutions (all without coding) for any requirement, the question no longer is ‘Can I solve this challenge?’, but will be ‘What is the best way to solve this challenge?’. The architect in me encourages this a lot. Do I add a formula field to the Opportunity object in Salesforce, do I add the formula field in the dataset definition or do I create the formula field in a compare table within the actual dashboard? Given a specific requirement, the answer to each question could be yes.
  • A dashboard can be embedded on a Lightning page or on a Community page. This is great, since now you can have the record detail information together with for instance a trend graph on one page!
  • Dashboards can be kept a lot cleaner by using links to other dashboards and dynamic filters to reduce the number of charts needed. More options and less distractions, that is something anyone would appreciate I think. Einstein Analytics has become very actionable. Do you want to navigate to a specific record from a Analytics table? Do you want to create a Chatter post based on a trend graph? Do you want to add a note to a Case? These actions, and many more, are available directly from a dashboard table in Einstein Analytics.

Action from a dashboard table in Einstein Analytics

Einstein Discovery

Another recent addition to Salesforce Analytics Cloud is Einstein Discovery. With Einstein Discovery every Salesforce user gets to be a data scientist, using both Salesforce data and external data. Einstein Discovery enables anyone to discover relevant facts and themes about their data, and it takes care of the difficult statistical calculations that help answer questions like ‘Why did it happen?’ or ‘What will happen?’. Those answers can help you take the correct decision, whether your data is about Sales, about Service, or about anything else. And in case you do want to access the statistical analysis it’s just one click away!

Einstein Discovery story example 

This Trailhead module describes the basics of Einstein Discovery very well. Setting up your datasets and your stories is very simple and intuitive. I also like the button with which you can generate a summary (that can be personalized as well) which enables a management presentation straight from Analytics Cloud.

While Einstein Discovery is a relatively new product, I predict that it will get very successful for companies that already use the Salesforce platform. Here’s why:

  • Adoption should be great as users are already familiar with Salesforce and with the data that exists in their Salesforce environment.
  • Unlike traditional statistical methods and tools in which a user had to specify which variables he wanted to be analyzed Einstein Discovery uses all data that is available! No longer do you need to ask ‘How is my average CSAT score impacted by channel and agent experience?’. Just create a story maximizing the average CSAT and Einstein Discovery will tell you what the significant variables are and how they impact the average CSAT. And those variables can come from anywhere in your Salesforce environment (or external data source), whether you would expect these relations to exist in the first place or not.

In this article I provided the main reasons why I think both Einstein Analytics and Einstein Discovery can be great additions to a Salesforce implementation. Please do not hesitate to contact me for additional questions or advice.