21 Apr Predictive Analytics: From Mad Men to Math Men
Today I have the pleasure of being at the Mid-Atlantic Marketing Summit. Hosted by Capitol Communicator and Potomac Tech Wire, the annual event brings together some of the top marketers in the Washington, D.C. area for a day-long event discussing all that’s hot in marketing. We’ll be sharing what we learned at today’s event through a series of blog posts over the next few weeks.
We shared our takeaways on the morning’s first two keynotes on Twitter and the first session I wanted to recap for our intrepid readers was led by Mitchell Eisen, CIO and co-founder, Real Magnet. Mitchell’s panel was titled “Predictive Analytics for Real World Marketing.”
Today you can’t go online without seeing and hearing terms like artificial intelligence, machine learning and predictive analytics (PA). There is so much hype and promise around these technologies, but do you really know what it means and where it’s going? Mitchell offered up the following definition of predictive analytics to kick us off:
Predictive analytics is the use of data, statistical algorithms, and machine learning techniques to identify the likelihood of future outcomes based on historical data.
You can probably already tell since they are underlined and bold that the two keywords in this definition are data and likelihood. There has been an explosion of data with the vast majority created in the last three years alone. At the same time, computers are getting smarter, faster enabling them to crunch all of this data to give us the insight to help predict what might happen. However, in order to succeed with predictive analytics, you need lots of data and you need it to be lots of varied data.
Mitchell warns though that as great as predictive analytics are – predictive analytics in marketing is more art than science. You are trying to predict human intent and predicting human intent is really hard to do.
Even companies like Netflix and Amazon – which spend tons of money on predictive analytics — don’t always get it right. Think about your Netflix or Amazon recommendations for shows to watch/items to buy. When was the last time you looked at them and thought “yep, that’s spot on!”?
So, with that, who’s using predictive analytics? According to a survey (that I’m so sorry I didn’t get the exact info on), 40% of responding marketing companies are using or intend to use predictive analytics in the next year. Out of that 40%, 53% responded that they’ve seen a significant boost to their sales as a result of PA. Less positive, only 7% feel that they are sufficiently staffed for predictive analytics. Technology is moving faster than human behavior and properly staffing your marketing team to make the most of predictive analytics should be a priority. Mitchell also points out that typically when using predictive analytics in the marketing realm, there are key areas that people focus on: 1. acquiring new logos/customers and 2. prioritizing/targeting existing customers.
When looking to acquire new logos or customers, predictive analytics can expand on the data already found in your CRM and offer insights on which target companies to focus on based on a variety of factors. It’s on you to capture the data but predictive analytics can enhance your data and offer additional insights. The one thing predictive analytics can’t do? Be creative. Computers are smart but once you identify your targets, be creative in how you reach out to that target.
One example of predictive analytics in practice is lead scoring. Lead scoring attributes a score to certain actions that leads/prospects take and can help you prioritize your efforts on who is likely to buy from you and at what point.
Just as important as understanding the customer journey, perhaps more so, is knowing when you are at risk of losing a customer. It’s much harder to acquire a new customer than to keep a customer and predictive analytics can help you keep an eye on churn. According to Mitchell, churn is a mission critical part of any marketing effort and predictive analytics can be your early warning system, letting you know if a customer is likely to churn and when. Being armed with this information allows you to take actions to keep that customer. This is something I’ve experienced personally recently. A brand that I frequently shop with noticed I hadn’t purchased anything in quite a while and emailed me with a personal offer of $20 off my next purchase. It certainly got my attention and while I haven’t used it yet I certainly intend to!
Wrapping up, Mitchell alluded to the hope, and maybe even promise, that with predictive analytics you’ll be able to target a person or type of person with the specific type of content at the specific moment to generate the outcome you desire. While predictive analytics and marketing automation is changing how marketing departments function, the need for smart marketing people has never been greater. Those marketing teams that can harness the combination of technology and humans will be the ones that succeed. Predictive analytics can’t do the work for you, but it can help you focus and optimize — turning your marketing department from Mad Men to Math Men.