Salesforce is introducing artificial intelligence (AI) and predictive capabilities across its cloud customer relationship management (CRM) software-as-a-service products through its Einstein brand.
At the time of launch, during the company's Dreamforce conference in 2016, Einstein general manager John Ball, said: "You will see [Einstein] powering AI across all of our clouds and there are many features coming. We do three releases a year and Einstein will be powering every release from now on, starting with the winter release in October."
Pricing for Einstein isn't clear cut, with some capabilities being included with existing cloud licenses and editions, and some coming at a premium. The pricing of each feature will only be known once it is made generally available, except Community Cloud, Commerce Cloud and Analytics Cloud capabilities, which are included in existing licenses and detailed below.
Salesforce has been working on Einstein for the last two years, spending more than $4 billion on AI specialist acquisitions like MetaMind, RelateIQ and BeyondCore, and building a team of 175 data scientists along the way.
What is Salesforce Einstein?
Ball said that Einstein is built into customer success platform "to surface those insights in context for business users".
"We enable sales professionals to find better prospects and close more deals through predictive lead scoring and automatic data capture to convert leads into opportunities and opportunities into deals."
AI is only as powerful as the data which powers it, but Salesforce has plenty of that, training Einstein predictive models on a range of data collected by Salesforce products. This includes: customer data, activity data from Chatter, email, calendar and ecommerce information, social data streams and even IoT signals. Einstein will continue to adapt to changing user behaviour as data comes into the cloud platform.
Salesforce is also opening up Einstein capabilities for App Cloud users and developers to bring AI features, such as predictive or suggested actions, into new or existing apps.
Ball said back at the time of launch: "Part of the vision and reality today is that Einstein is baked deep into the customer success platform, so taking and configuring and extending existing apps with Einstein fields and partners building on the customer success platform means Einstein will make all apps smarter, including those on the exchange."
Salesforce Einstein features
Salesforce promised that it would be infusing all of its popular software-as-a-service (SaaS) platforms with intelligent Einstein features when the brand was announced at Dreamforce in 2016. Now, over one year on, customers are starting to see Salesforce roll these features out.
Then at Dreamforce 2017 the company announced two new capabilities to make it easier for admins to build AI-powered apps and chatbots within Salesforce without writing a line of code, called MyEinstein.
First there is Einstein Prediction Builder, which allows customers to create custom AI models around predictions for any field that is currently held in Salesforce simply by identifying the field they want to predict and selecting the data they want to use with a simple point and click interface.
For example, someone in the finance team could create an app that leverages the data in Salesforce and a pre-packaged machine learning model to predict which customers will file their invoices late. Einstein Prediction Builder is currently in pilot and expected to arrive in summer 2018.
Then there is Einstein Bots, where customers can create chatbots powered by historical service and CRM data to respond to common customer inquiries and deploy it through Service Cloud.
For example, a bot could be set up to track order status or request a refund and trigger the relevant process automatically. Einstein Bots are currently in pilot and expected to be generally available in summer 2018.
Now, here is a quick cloud-by-cloud breakdown of the AI-rich features which have been announced so far.
Salesforce announced three features for its Sales Cloud Einstein product in September 2017, due for general availability in early 2018.
The first is called Einstein Forecasting, an out-of-the-box tool for sales staff to make accurate forecasts using all of their historic CRM data. Einstein Forecasting is essentially a bundle of self-learning algorithms that learn individual and team forecasting behaviours to offer objective insights into future sales.
Salesforce also announced that Sales Cloud customers will soon be able to leverage Opportunity Scoring to automatically prioritise high-value opportunities and Email Insights, which uses natural language processing to identify the most important emails. The pricing of these Einstein features will be announced at the time of general availability.
Salesforce announced a range of Commerce Cloud Einstein features back in May 2017.
The first new feature is called Predictive Sort, which uses machine learning to personalise the order in which products appear in search and category pages on ecommerce sites, down to the individual shopper depending on their previous browsing habits.
The next feature is New Order Management, which allows retailers to connect customer demand with inventory supply by analysing order and inventory data across stores, warehouses and dropship vendors to support “buy anywhere, fulfill anywhere” scenarios.
These features join the Einstein Product Recommendations tool, which recommends the best products to a consumer across various channels.
Service Cloud Einstein will include recommended case classification which will automatically pre-populate key case fields and route them to the right agent predictively. Recommended responses will push the most likely responses to service agents and the platform will predict close times related to issues.
Marketing Cloud Einstein now includes predictive scoring related to the likelihood of a customer engaging with an email, automated send-time optimisation and predictive audiences, which will build custom audience segments based on predicted behaviours.
Salesforce also added an AI-powered image recognition tool, called Einstein Vision for Social Studio, to its Marketing Cloud platform in August 2017. This allows marketers to automate the discovery and identification of images shared on social media, even when they haven't been mentioned specifically, significantly speeding up response times.
Community Cloud Einstein now includes recommended experts, articles and topics, automated service escalation and newsfeed insights. Automated community case escalation and recommended experts, files and groups are generally available and included as part of the Community Cloud license.
Analytics Cloud Einstein customers will be able to access predictive wave apps to uncover future patterns for any business process, smart data discovery which will help users find and explain insights from their data, and automated analytics to prioritise insights. Smart data discovery is generally available and priced according to the volume of data and number of users.
IoT Cloud Einstein will include predictive device scoring, recommend best next actions for service processes and marketing journeys, and automated IoT rules optimisation.
The company's vertical-specific tool Financial Services Cloud also got its own range of Einstein features in March 2017.
Opportunity Insights gives financial advisors the ability to uncover opportunities based on client sentiments, competitor mentions and overall engagement by analysing previous email conversations.
The Relationship Groups feature helps advisors link clients to multiple households, trusts and business groups to ensure they have an up-to-date view of their client's wealth.
Finally, the Relationship Map gives advisors a visualisation of a client's family wealth ecosystem and financial accounts.
Find your next job with computerworld UK jobs