In the post-Cambridge-Analytica-and-GDPR-compliance business landscape, there have been more than a few lengthy, awkward boardroom silences when the topics of customer data analysis and targeted personalization have been raised. It's ironic that just as the computational and academic horsepower to perform apparent data wizardry have become mainstream, accessible realities for many organizations, the fear and uncertainty around doing so are at an all-time high.
These days, organizations are quite correctly focused on the protection and informed, consented use of personally identifiable information (PII). For that reason, nobody should be shipping off reams of addresses and social security numbers to their analytics or data marketing partners. Conversely, though, we ought not miss out on the opportunity to win big with an ROI-focused data strategy that still colors well within the lines.
Non-PII traits are the goldmine of insight into who your customers really are and what factors drive their behavior. Focus on building connected datasets that contain as much of this information as you can get your hands on—void of anything directly identifiable, of course— and bind them to an internal customer ID so they can be actionable once you've formed a hypothesis. From there, get it into the hands of your analysts and let them have at it.
Segmentation isn't what it used to be
Traditional segmentation tactics often use a single variable as a proxy for a collection of traits that are much more nuanced. For example, you may assume a younger demographic behaves a certain way or will respond to a particular message due primarily to their age. Without also considering the interplay of other highly influential variables such as income, education, occupation and cumulative behavioral signals, you may be making a big mistake. While this may have made practical sense for early database marketers, given their technical constraints, today's capabilities allow for a mind-boggling array of statistical methods that will reveal complex profiles that represent huge value potential. Just be sure not to slice too finely with your segments or you'll risk losing impact as people start feeling uncomfortable about exactly how specific and detailed your messaging is.
Previous performance indicates future behavior
Every customer who ever stopped being one, for any reason, left a trail of clues as to why they did so, and it's likely their name or address has nothing to do with it. Go looking for the patterns hidden within your existing audience and customer data: login frequencies, session durations, wait times, customer support inquiries, review sentiment, purchase frequency. The list goes on. Predictive models of negative outcomes can play a key role in a well-executed retention and communications strategy that will drive measurable revenue as you begin to intervene at critical points, causing your churn rates to diminish.
It's the little things that make personalization personal
An authentic personalized experience for your customers need not infringe upon their rights to privacy or exploit their personal biases in a manipulative way. Audience management tools and data management platforms (DMPs) are getting some serious and well-deserved attention these days. In-the-moment activation of the things you know about your customers and what they are looking for are very achievable tactics and should feature predominantly on any data activation roadmap.
Ask yourself these questions:
- Who chose the current hero image or primary call to action on your homepage?
- Why was it chosen?
It's likely you don't have terrific answers right now. Don't worry. You're certainly not alone. But imagine you could respond with:
- "My customer did!"
- "The anticipated intention or preference for their visit."
Something as simple as a subtle shift in imagery to align with gender, product interest or personal affinity can often yield a profoundly positive impact on your bottom line.
Today's savvy data marketers arm themselves with the right blend of strategy, technical skills and a determination to overcome prevailing methods, while pragmatically addressing privacy concerns. There are many data strategy opportunities available that don't creep out your customers or endanger losing the hard-earned trust they have placed in you.