March 16, 2021
By Max Braun, Associate Director, Experience Planning
Pick a color, absolutely any color — as long as it’s blue, yellow, or green.
Free will is tricky like that, especially in a reality where choice is king but options are limited by technology and data availability. Data-driven marketing is partially to blame, as is our newfound preference for engaging with only companies and content we align with. Advertisers love to talk about empowering customers with the freedom to choose, but brand interactions tend to have the opposite effect.
It’s an impossible balance to strike. There’s a tension between anticipating what an individual actually wants without needing their input versus letting the customer indicate their own preferences. As a society, we’ve become so jaded by companies like Google and Facebook, whose algorithms silently deliver against our expectations on what it means to be predictive — while at the same time eroding our trust that they have our best interest in mind.
There’s an inherent problem with comparing your ambitions to the technology of Facebook: The average adult spends roughly 38 minutes a day on Facebook, and many have been active on the platform for over a decade. This data disparity means that Facebook has trillions of data points against years of activity to predict what every individual wants to see.
To be sure, marketers can’t set our ambition to achieve the kind of predictive intelligence that Facebook has — at least not right away. We need to set our ambition against the time and attention the individual actually gives to our business while considering their historical relationship with the brand or product.
Many brands recognize this challenge and try to give the customer the ability to build a profile of communication preferences. One problem is that many brands assume that a customer wants to receive every triggered communication and interaction until they specify otherwise, prioritizing a sales-based approach through every point in the relationship.
No, the best strategy balances artificial intelligence and predictive analytics to give customers what we believe they want while also giving individuals the freedom to adjust their own levers. In the absence of good data and information about a customer, you can and should leverage their own inputs to build a progressive profile about what they want to see versus making assumptions based on categorizations that may incorrectly generalize an individual.
The Perfect Intersection
The real problem here is when we use incorrect, incomplete, or assumptive data to inform what a customer sees. We often are afraid to show the inner-piping of our business, assuming that too much choice is problematic and destructive. And though it is true that too much freedom can lead to customer inactivity and stagnation, the right balance should be to define a road map for the individual to get to pure predictive content and communications (without them even realizing it).
Rinse, a dry-cleaning app, mixes its sales-based and inactivity text messages with its appointment confirmation messages. It has defined a period of lapse against the “average visit rate” and begins sending sales messages as if a user is inactive when “most other people” are inactive. The problem here is that one’s inactivity is not necessarily due to lapse but rather a reflection of lower usage patterns, which could be environmental or unrelated to the relationship with a business itself. A rigid communications pattern, even if it’s against a behavioral model, leads to a bad experience. A way to combat this with appropriate consumer choice is to ask, “How often do you typically use a dry-cleaning service?” It’s a simple question that could personalize message cadence with sales messages at the right time and place.
Every individual’s needs are different, and while we often like to start a relationship off on the right foot, too often we pull media segments through to that first interaction without allowing a customer to personalize their own experience. Without going further than demographics and search behaviors, we probably don’t know enough about an individual to deliver the correct, relevant message at onboarding. Don’t underestimate the power of letting the customer provide input.
Inspiring Customer Control
You can get to a more seamless transition to powerful and accurate predictive content in a few ways:
- Inquire. You can ask the customer to provide their thoughts at the beginning of a new relationship on why they’re there. Brand perception certainly plays a role in these decisions. Nurturing positive customer experiences is the best referral money can buy, after all.
- Interpret. Consistently check in with the customer. Too much prodding and poking can be annoying, but if you’re asking the individual how we’re doing and how they’re doing, it’ll demonstrate a commitment to care and give them control.
- Input. Introduce predictive controls over time versus using a light switch. The best implementation of a behavioral model adjusts with reactive behavior over time rather than forcing customers to fit into a box.
Predictive content should be the North Star for any business looking to get more anticipatory personalized interactions without annoying our customers. But appending it in the right way, at the right time, and over the right period is integral to delivering a relevant experience. Even the best predictive models can be making inaccurate assumptions about the audience if it’s not appended to content in the right way.
By letting the individual take control over their own destiny, we not only give them power, but we also gain insight into what they care about most.