Today's blog post was written by Kristie Reid, VP of Consulting at Sonoma Partners.
The latest CRM buzz is around Artificial Intelligence. No, not the kind of AI that allows computers to take over the world (yet) or the latest winner of Jeopardy. I mean the kind of machine learning that can process your CRM data to reveal some pretty amazing things. Some of the use cases thought of so far include:
- Predictive analytics for lead scoring
- Proactive notification to identify deals going south before the sales person recognizes what is happening
- Understanding customer sentiment without needing to pick up a phone to see if they are happy or dissatisfied
- Automated email creation that inserts the content before you can even think of what to write
Salesforce made a huge announcement a few weeks before Dreamforce about “Einstein.”
This is what they are coining as their machine learning technology built directly into the
Salesforce platform. In true Salesforce fashion, they went big at their annual conference with adorable Einsteins running around with the 170,000 attendees.
Microsoft doesn’t have a catchy title like Salesforce (or a cute logo), but they do have Azure Machine Learning. This product is currently more of a platform which can be configured and incorporated into your Microsoft products, including Dynamics CRM (or Dynamics 365 for Sales these days).
No matter which product you use, this is exciting stuff for CRM applications which have historically been thought of as overhead. Imagine telling your sales team that CRM can now write their emails for them!
But before announcing how smart your CRM system is to your organization, here are some things you will want to consider:
- You must feed the beast: The predictive analytics engines that power these tools require data for them to analyze. So, for new CRM implementations, this may take a while. Be realistic about what you can expect and when.
- Garbage in makes even worse garbage out: Artificial Intelligence does not resolve the age old issue of "bad data in, bad data out." Except now, there may be more risks exposed since "bad data in" could lead to bad decisions.
- Who’s right, who’s wrong: What if the output from the machine learning algorithms doesn’t match your rules? These instances can be taken case by case, but this is something that should be monitored by a business sponsor who understands.
Questions/comments/concerns? Give us a shout.