I’ve been in this game for a long time. (OK, I’m old). But there is one old trick that never seems to wear out. If you want a really fast return on investment for something innovative, just take two trends that are bringing value and figure out how to put them together. Today’s case in point? Marketing automation and machine learning.
Every B2B company I know is pulling all the stops out in marketing automation. They are creating offers targeted to their key segments and they are using analytics to choose the winners and their roadmap for 2015 shows how they integrate MAT (that’s cool-talk for marketing automation) with their CRM system (that’s cool-talk for, uh, a CRM system). In other words, marketing automation is one of the flavors of the month. All the cool kids are doing it.
Now, one of the problems is that many of the cool kids are doing it wrong. They are mostly using marketing automation for outbound marketing, which, while interesting, is not the only game in town. The even cooler kids are using it to tune any message (outbound or inbound) to specific micro-segments. In other words, they are using machine learning techniques to analyze thousands of data points to decide which offer works with a person like this one.
This is not your father’s A/B testing.
I spoke recently with John Bara, President and CMO of Mintigo, which he describes as being in the (buzzword alert) “enterprise predictive marketing space.” Because John is smarter than me, I had to ask him what that meant, and he helpfully replied that “Predictive marketing uses data to make machine-based recommendations on which actions convert leads to customers.”
Folks, what he is saying is that machine learning has met marketing automation.
Now remember the theme today–put two hot trends together and get something that most people would call innovative. But just doing that yields hype rather than ROI. So, to make sure that you aren’t hearing a story about where the rubber meets the sky, you have to dig deeper. Mintigo has over 4,000 data points (oh no, Big Data, too?) on every person you are targeting.
But does it really work? RedHat says it does. They used Mintigo to figure out that they should be targeting Jira users. Now maybe to you it seems obvious that a company that services the Linux operating system should hang out where coders do, but the point is that it wasn’t obvious to them. You probably have similar blind spots.
Any approach that adds machine learning to marketing automation has great potential. You can examine a set of leads with their outcome data–that’s machine learning talk for whether the leads closed, were lost, or are still in play. Then the machine learning algorithm identifies patterns in the data to help you identify the characteristics of your best prospects–the ones most likely to close.
Why do you care? Because if you are like most companies, you have more leads than you can service with your inside sales team. So, wouldn’t it be great to see the best leads get routed directly to your top salespeople? That’s what machine learning can do for you. It can identify the best offer for different types of people.
What trends are you watching? And which two can you put together to really supercharge your results?
