Why Retail is using data the wrong way – ensuring the best customer experience for omni retail

mattress selection and computers

Today a sales associate tried to help sell my mother on a mattress through their special diagnostic tool.  They said that it compiles thousands of data points from sleep scientists, and uses thousands of sensors to guarantee the perfect fit.  This pitch is nothing new.  This approached is used in dating, shoe selection, and multiple verticals for the consumer space.  However their are several major issues:

1.  It assumes everyone is the same.  Statistically, a computer can compile that most people will like the same thing, but we all have personal preferences as well.   The take away should be that many people feel this way as a guide, rather then that this is the “right” fit for you.

2. Salespeople still have to treat people like people, not equations.  If the sales associate would have talked to my mother before throwing her in a machine, we would have been more likely to buy a bed.  Data is no substitution for customer service.

3. In Certain Data sets, Data only gets you so far. In many situations, tests can get you part of the way there, but after filtering out some basic traits, the rest has to be decided in person.  For example, an online search can tell me which cars might be a good fit for me based on price and preferences, but I still want to test drive them to see how they feel.   The issue is that retail associates mistake tests for absolute answers, rather then tools to help move someone down a funnel.

4. The Bell Curve is an outdated model in many verticals.  In our culture of abundance, we often look to machines and software to tell us the most efficient answer.  For verticals that are based on taste, the bell curve model is dead.  Content consumption ( videos, music, etc. ) is a great example where people can find the exact fit for them, rather then looking for the most popular item.  Digital goods is the easiest argument for this last fit. However, other physical goods cater to clusters of people or outliers in the equation – take Etsy for example.  It caters to the long tail of economics, finding custom goods for custom buyers.    Now for commodity good like a bed, the bell curve still may be a good fit.  This may speak to the cost and scale of production, but I imagine that the availability of options in the mattress industry is changing consumer expectations as well.


In any instance, as retail has the ability to be more digital, it’s important to maintain some elements of care and humanity. Tools can mutually help people be more efficient with sales and customer satisfaction, but its important that tools are used the right way to help out both parties, with a grain of salt for personal preferences and individuality.

As an overall generalization for omni channel experiences,  data and software is great for helping people work through areas of the upper funnel for customers, but often the later parts of retail need a high touch experience to fully ensure customer product fit and maximum satisfaction.  Both digital and analog need each other for certain verticals to be truly successful.