Einstein Discovery allows users to gain insight into their data quickly and make better-informed decisions regarding their business. While very complex data questions may still require the assistance of a data scientist, Einstein Discovery puts the power of analytics in the hands of those individuals who own the data — allowing you to make smart decisions on the go without needing to create complex models. Let's dive in!
So you've decided that you are ready to use the different types of analytic frameworks to start asking the right questions about your data. You're off to a great start!
But digging in can be intimidating. How do you even begin to analyze all of your data when there's so much of it? Salesforce has the answer with Einstein Discovery.
Einstein Discovery is Salesforce's AI-powered analytics tool that will help enable business users, like yourself, to discover relevant patterns and trends quickly.
With the ability to analyze millions of rows of data within minutes, Einstein Discovery allows business users to ask critical questions regarding specific KPI's to inform you of:
- What happened;
- Why it happened, and
- What could happen
Einstein Discovery not only analyzes your Salesforce data but can also process data from external systems and spreadsheets, reducing the need to toggle between multiple systems or sources. Have a KPI that you'd like to maximize or minimize? Want to know which leads are most likely to convert?
No need to search through thousands of rows of data across several platforms, ask Einstein!
Understanding What Happened
Following the Descriptive Analytics Framework, Einstein Discovery analyzes your existing historical data to determine what has happened in the past related to your KPI in question (known as your "Outcome") and displays the results in a "Story." The Story contains graphs and descriptions that inform you of the patterns in your data that the statistics show were most influential on your Outcome. The Story also highlights any information within your data that has an essential impact on your KPI.
For example, let's say you asked Einstein: "Hey Einstein, how can I maximize my conversion of leads?"
Einstein would analyze all records related to your leads and inform you of all variables, and variable combinations, that it has determined will result in the successful conversion of leads. After creating your Story, you have found that all Leads obtained via your website have a 75% chance of successfully converting. Great! But why is this the case?
Understanding Why It Happened
Now that you know what has happened within your data, your next question is probably, "Well, why did that happen?"
Based on the patterns and trends found during its analysis, Einstein Discovery will provide you with the specific factors in your data that have resulted in the Outcome. Using interactive waterfall charts, Discovery allows you to dive further into your analysis by allowing you to see how certain variables or combinations impact your overall Outcome.
Using our previous example, Einstein Discovery will allow you to explore further into your data to see that not only are Leads from the website successfully converting but, of that 75 %, 70% have an Account Type of "Distributor" and 2% have an Account Industry of "Retail." Great! Now we know that our Distributors are using our website and successfully converting while our Retail division doesn't appear to be converting as successfully. Now we know why our conversion is high, but how do we make that even higher?
Predicting What Could Happen
Following the Predictive Analytics Framework, Einstein Discovery allows business users to change variables within their data to determine the future impact on the Outcome. Based on the requested changes made, Einstein Discovery will provide users with any top improvements to be made based on the desired change as well as both positive and negative factors of making the change. Lastly, Einstein can provide you with recommendations of changes to make to your data that can help to influence your desired Outcome.
Using our lead example again, we want to know what we can do as a business to increase the successful conversion of retail leads from our website. Utilizing the "What Could Happen" insight, we can make subtle changes to different variables in our data. We might look at the company size, lead source, lead rating, etc. to determine how these changes can either increase or decrease our lead conversion rates. By changing some of our variables, we can learn that by targeting retail companies with a size of 100+ employees, we can increase our conversion among retail industry leads from the website from 2% to 15%. That's great news! Time to work with the marketing department to better determine how to target these companies.
What are some relevant examples you currently use Einstein for? Let us know in the comments below!