What’s Still Missing From Hyperpersonalization Strategies

What’s still missing from hyperpersonalization strategies? Context, situation and situational markets. In fact, "hyperpersonalization" is a problematic idea. At least for retail. Let us explain why.

People Want Customized Retail Experiences

Retail experiences that are customized to individuals are an absolute must today. People will not tolerate retailers that don't remember purchases, provide tailored recommendations, and understand preferences. Many retailers refer to this kind of activity as personalization.

In most cases people want more ability to customize their experiences and some companies have referred to this as hyperpersonalization. But when we talk to consumers, we hear something very different.

What they want is the ability to control their customizations. In other words, they think they know better than the company’s algorithms what they want from customizations. They don’t want the company to do the personalization for them. They want their hands on the levers.

For this reason, we often call the type of customization that people want today "individualization" rather than "personalization." Individuals want control. Companies that retain control may personalize the experience, but they must guess what new levels of customization are required.

The Algorithms’ Errors

The first issue with hyperpersonalization is that what people want is more control, not more personal recommendations. The second issue is that most of the tools that are designed to customize the purchase experience for consumers use the wrong data to predict future needs.

Predictive algorithms use Bayesian mathematics to anticipate what the customer is likely to want. The math is based on the principle that past purchases are the best predictor of future purchases. Seems obvious. But think about it: if you just bought new tires for your car, does that mean that the next purchase you will want to make is an oil change? What consumers experience when they return to purchase from a site is a form of déjà vu: “the reason why they are offering me air fresheners is because I bought tires.”

But what the retailer can’t see, and the algorithm can’t anticipate, is that the consumer’s situation has changed. Their reason for showing up to buy is focused on something completely different. It may be because they need to buy a birthday present or supplies for a football game.

Thus, the thing that drives the purchase is the situation, not the preference. If companies want to tailor their purchase experience to the individual, they should focus on common situations. Then they need to build their algorithms to recognize those situations and make recommendations based on what others wanted when presented with the same situation.

Situation-based analytics will beat hyperpersonalization analytics every time.

Dave Norton, Ph.D. is the founder and principal of Stone Mantel, a research-led consultancy at the forefront of customer and employee experience strategy.