Can a Customer Be a Number?
Analytics gives you an imperfect view of your customers.
If you liked reading this, please click the ❤️ button on this post so more people can discover it on Substack. Thanks!
One of the great innovations in business in the past 50 years has been the rise of analytics. Using data to make decisions has improved efficiency, productivity and growth across all industries from the smallest companies to the largest.
Today, it’s rare to find a company that doesn’t use analytics as it’s primary view into the health of the business. Since businesses are built on customers, that means using analytics to understand customers, whether they are happy and how they are buying. Regardless of whether you sell ads, couches or software you are watching your customers using analytics.
As someone who has spent most of his career in the analytics field, I think this is great! It’s certainly an improvement over the days of using intuition or opinions to make major decisions.
At the same time, I fear that something has been lost in our relationship with customers. Customers are people and people are poorly understood through analytics alone.
For example, analytics can tell us what customers have done but not why they did it. We can try to use A/B tests and other methods to try and determine their intent and motivation, but even then we are guessing. This is partly because data is a historical record of what happened in the past, but it’s mostly because people are complicated.
People are Complicated
One of the underlying assumptions of analytics is that the customers being analyzed are making rational decisions. If you look at the traffic activity on your website, you assume people are not clicking around at random. You assume there is some method to their madness, and you are trying to use the data to discern it.
But what if they are clicking on things at random? Or, at least, clicking on them because they don’t understand? How would you know the difference?
To get these answers, it’s tempting to think that talking directly to customers is better than using analytics. Afterall, what is better than asking a customer what they are thinking, and why they are buying? There should be no better way to get straight answers than straight from the source.
Unfortunately, if you were to interview a dozen customers about why they bought something from your business, you might get a dozen answers. If you interviewed them again, you might get a dozen different answers from the same customers! Most of us poorly understand why we make decisions in the first place, but have an even worse memory of decisions we’ve made in the past.
If customers can’t tell us their intent and motivation in making a decision, we can’t expect analytics to do any better. When you think about it, the flaws in using analytics to understand customers are because it’s nearly impossible to understand customers in the first place!
Right, so what?
Should we stop using analytics because they are imperfect? Of course not! Analytics will always be a key part of how we run our businesses. The better you use analytics the more efficient and productive your business will become.
We just have to be very aware of the limits of analytics, and specifically the imperfect picture of customers that analytics provides. We need to stay humble and realize we are looking through a foggy window, trying to make out sharp details that are impossible to discern.
Analytics are our view into business, and while imperfect, it’s the best one we have.
For more on Analytics, see:
There is a difference between Success & Velocity metrics
Always choose Second Order Metrics
As someone who has had some data analytics companies as clients, I must say Sean's point of view blew me away. Yes, customers are humans, after all, but we've reduced them to cold statistics. A thought-provoking piece.
good article