If something bad happens to you on Valentine’s Day, and malaise sets in for a period of time afterwards, that’s a Valentine’s Hangover.
A similar feeling may set in for marketers embarked on their first Big Data projects. There’s a lot of passion and excitement leading up to it, only to be quelled by unsatisfactory results, or worse, learning that the integrity or usefulness of the data isn’t worth a lick. And if the project is a complete train wreck, there’s no love lost between IT and marketing.
There is a preventative measure. Hangovers typically result by starting too quickly, running wild and free, or mixing too many different “variables.” So rather than make your foray into Big Data with sexy projects like dynamic pricing, predictive analytics or machine learning, consider basic data mining as a first date.
Everything In Moderation
Data mining is a smart entry point because it’s an asset you already have. You’re not going in completely blind! Even the least data-savvy brands have leads, somewhere, in a database — perhaps from e-Commerce transactions or online registrations. And data mining is a technique that can help you uncover consistent patterns and relationships between variables, so you can achieve a better understanding of customer value (and what turns customers on).
State Your Intentions
Why are you doing this? What do you want to learn? Before you begin, collaborate with your partner team to clarify objectives. Some thought starters:
If you don’t know where you’re going, any road will get you there. So be clear about the endgame.
As with any personal relationship, it’s not always how much your love spends on you. The same holds true with customers. How frequently or how recently they’ve engaged might be more important to your long-term relationship. RFM analysis (recency, frequency, monetary) is a simple framework that quantitatively identifies which customers are the most valuable. Looking at how recently a customer bought, how often they buy, and how much they spend per purchase enables clearer segmentation. There are plenty of models out there. Choose one that provides the best match for your objectives.
Keep Logistics Simple
Agree to some steps that will help keep the engagement on track. For a data-mining project, it might look something like this:
Stop Trying To Prove Yourself
Don’t be surprised if your findings yield something completely unexpected. Be open to what you might find. And if you do find something wonderfully unexpected, be delighted! Otherwise, you’ll fall prey to look only for support for something you already had in mind.
Don’t Ignore The Awkward Moments (During The Presentation)
Data presentations tend to have lots of rabbit holes, largely because the cross-functional team you present to (executive level, marketing, IT, sales, finance) have different interests. And the reality is they’re new to this stuff, too, and will have a lot of questions. Questions are good, but too many questions kill the spark and spoil a good date. So make sure your presentation leads everyone down a clear path toward the most important insights and actionable findings. And if a significant objection is raised, or if something really off the wall is asked, be sure to frame it in the context of the objectives.
Dump The Three-Day Call Back
If you make a great presentation and your recommendations have cleared next steps, be sure to ask for your offer at the meeting. Don’t wait. Keep the momentum. Because to operationalize data analytics in your marketing program, you’ll need an ongoing commitment.