Not long after the closing bell rang on Wall Street yesterday, Salesforce CEO Mark Benioff and IBM Chairman, President and CEO Ginni Rometty announced a global strategic partnership between their companies to advance the adoption of artificial intelligence (AI) in the enterprise through an integration of Salesforce Einstein and IBM Watson.
The news came just hours before Salesforce announced the general availability of Einstein AI to all of its customers across sales, service, marketing and commerce as part of Spring '17, Salesforce's 52nd major product release in 18 years.
Yesterday's sales pitch to investors, customers and the media sounded something like this: by combining IBM Watson's capabilities to glean insights from structured and unstructured data across many industries (including weather, healthcare, financial services, retail) with Salesforce's CRM data, businesses — specifically salespeople, service providers and marketers — will be able to make smarter decisions faster.
Watson + Einstein = Greater Intelligence
Take, for example, an auto insurance company that wants to keep customer claims to a minimum. If the company knows there is a hail storm approaching (from insights gleaned from IBM's Weather Company), it can leverage the Salesforce CRM data of its customers in the affected area to prompt them to move their automobiles indoors. This sort of solution is expected to be delivered via a service that is, for now, called IBM Weather Insights for Salesforce.
There is a slew of other products around the partnership that already have names such as the IBM Watson and Salesforce Einstein Integration, which promises to bring Watson insights from shopping data, for example, into the Salesforce platform.
There is also the IBM Application Integration Suite for Salesforce, which financial advisors might use to bring insights gleaned from market data into their personalized customer presentations and so on.
The potential wins for Salesforce and its customers are obvious. What has not been disclosed is how much Salesforce (and its customers) will be paying for them. In response to a question, CMSWire was told only, "Pricing will be announced at the time of general availability."
Revenue Generating, Costs Unknown
IBM, for its part, has promised to adopt Salesforce's Service Cloud companywide, which will not only provide income for Salesforce but also a top-shelf, reference customer. Constellation Research founder and lead analyst R "Ray" Wang said this is significant because the Service Cloud is the fastest growing part of Salesforce's business.
The deal also introduces a new revenue-generating opportunity for IBM through Salesforce partner and service provider Bluewolf, which IBM acquired last year. This could potentially bring in earnings IBM desperately needs at a time when it is showing 19 consecutive quarters of losses.
Watson — arguably best known for beating Ken Jennings on Jeopardy in 2011 — has shown great promise before. The question is whether it can generate meaningful, reliable results fast enough. In fact, last November, IBM's much-heralded moonshot project with the University of Texas MD Anderson Center — a project that was supposed to "eradicate" cancer — was been put on hold because there was not enough to show for the $62 million already spent.
While Watson has some successful individual deployments, we have yet to see out-of-the-box scale across thousands of Software-as-a-Service (SaaS) customers, said Constellation Research VP and Principal Analyst Holger Mueller. Should that happen, it would be a game changer, he suggested.
Or maybe not. Mueller told CMSWire IBM has made it clear Watson will keep running on IBM, leaving questions open from a data perspective. Specifically: "How does the data get to IBM and will Salesforce customers allow the data to go there?"
Wang told CMSWire there are a number of notable wins for IBM across the deal, not the least of which is Salesforce's need to access massive compute power. Mueller explained that Salesforce was supposed to get this from Amazon Web Services (AWS), but AWS is "rebooting" its machine learning capabilities so this capability is not available.
Finally, it is hard not to wonder if this is a preemptive play against Microsoft, especially when you consider Microsoft has all of the compute power it needs on Azure, as well as access to world-class data scientists and machine learning experts that are at least equal to IBM's and all the data in LinkedIn, which Benioff so desperately wanted to own.