AI is showing up everywhere in CRM -- in increasingly creative ways -- because the cloud model makes it easier...
to deploy and cross-pollinate AI tools. 2018 brought many promising new innovations that will lead to even more AI proliferation in 2019.
One of the most prominent innovators is Salesforce -- bound and determined to see its Einstein AI alongside IBM's Watson in the AI heavens. Two new Service Cloud Einstein capabilities -- Einstein Bot and Einstein's Next Best Action feature -- came into the limelight this year and have potential to play out impressively.
Bots, Voice and the future of apps
Einstein Bots for Service Cloud is intriguing for a number of reasons, not least because it is codeless. Not only can a bot be built, tested and deployed to a production application with nothing but clicks, but it can also be configured to pass the data it gathers on to separate applications in Service Cloud. Einstein and bots can be contextualized to pivot to new dialogues based on key phrases and words with that same click-based utility.
It gets more impressive still with the addition of Einstein Voice, which will enable voice input into Salesforce apps through digital assistants, such as Amazon Alexa and Google Assistant, with Einstein then retrieving the relevant record referenced in the dialogue. Einstein Voice is being piloted in 2019.
Service apps may be the most natural and popular domain for bots, especially smart ones, because many simple service tasks lend to straightforward automations, examples of which include tracking packages, checking bank balances and changing passwords. Einstein AI will enable exceptionally good service bots, and many will spring up over the coming year. But this combination of platform integration, features and user utility foreshadows much bigger things.
For example, some fast-food restaurants have a self-service kiosk where you create your order and pay for it by tapping through a visual interface. These can be clunky and cumbersome and take longer than just speaking to the person at the counter.
Something like this could be a job for Einstein Bot. Users could give orders to a bot in a fraction of the time it can take to scroll through kiosk screens, achieving the same intended end result, while eliminating the need for a human order-taker.
In Service Cloud Einstein Bot applications along these lines -- where the bot becomes an app -- the crowning touch is that dialogue-driven app events can trigger workflows, and those workflows can be dynamic.
Einstein Next Best Action and Bot workflow
If Einstein Bot is hinting at a new voice-driven future for apps, the Next Best Action feature may be signaling a shift in the future of workflow.
Next Best Action -- a big data recommendation engine that applies predictive analytics to business rules -- is intended primarily as a dynamic workflow tool for salespeople to keep the pipeline flowing smoothly. It's a course correction mechanism for sales operations in constantly changing circumstances.
Available to select Salesforce Einstein AI customers in a pilot program that began over the summer, Next Best Action is of obvious value in its parade application -- dynamic customer cultivation. But that's the tip of the proverbial iceberg in terms of its potential. More broadly applied -- and combined with Einstein Bot -- it opens up entirely new possibilities in interactive apps.
Imagine, for instance, that Next Best Action is not rendering its predictive recommendations to a human salesperson, but to Einstein Bot. And assume the bot is not a service app but instead lives in human resources. And its function is to help employees choose health plans.
In such a scenario, employees would respond to prompts from the bot to describe their health care needs and preferred options, working through details like optimum co-pay, deductible and so on. Einstein Bot's ability to contextualize would serve up appropriate changes in dialogue, based on key phrases spoken by the employee. The handoff of collected data would not simply yield a decision tree recommendation but could trigger Next Best Action, which would offer a more robust predictive outcome of plan selection, in turn, shifting the bot's dialogue to help guide the employee to the best possible choice.
Salesforce has the tools, then, for creating dynamic, voice-driven apps that can tap big data for real-time predictive analytics, which can, in turn, improve the user-bot dialogue. The resulting enhanced applications will be far more robust and easy to use than their web app predecessors -- and all without a single line of code.
Whether or not users end up exploiting these new Einstein tools this energetically remains to be seen. But the groundwork has been laid. Whatever we see from Service Cloud Einstein in the coming year, it's going to be interesting.