insight

When data analytics go wrong

Not only can they go wrong, but they can end up harming your business. Just ask Facebook.

Mark Zuckerberg has had a bad year.

After Trump’s election, many pointing fingers blamed Facebook for serving fake news, growing filter bubbles, and even being a platform for a hate speech. Thank you and goodbye!

Obviously, people are spending less and less time on Facebook and asking why are they using the service or what is the value they get from it?

Facebook’s downturn and escaping users show how data-driven thinking can badly mislead us. Facebook has among the best data analysis capabilities in the world, and it still can’t even understand what is good for its users.

Of course, Facebook has given us unparalleled opportunities for interaction with each other, but on the other hand, it has sadly evolved into a platform of addiction. If we consume more and more of our time by staring at our screens, it doesn’t make us happier or healthier. Facebook has learned many data-driven tricks to get most of our attention and to trigger a dopamine rush in our brains.

A data-based quantitative study shows us how things are in general, but usually, it doesn’t tell us why they are that specific way.

Dopamine makes us feel good in the same way winning a game or getting a hug does, but over a long period, Facebook’s triggered dopamine only makes us want more. Happiness is easy to confuse with pleasure and even easier if you’re already miserable. Any hedonistic pleasure could feel good for a moment, but it vanishes as fast as it emerges and leaves us alone with our desires, daily to-do-lists and guiltiness (if we are too addicted to sudden fulfilments).

To overstate matters a bit: Facebook has made a simple assumption based on data analysis: because we’ve collectively spent more and more time on the service, then more Facebook is what we need.  

The more data we have, the more we need to ask

For me, it seems like Mark is suffering a common disease of this era: a data-hubris, an over-confidence. Data gives us confidence, deservedly so, but this is also dangerous if we start to rely on quantitative analysis too much. Or even worse: if we let raw data lead our businesses instead of human thinking.

A data-based quantitative study shows us how things are in general, but usually, it doesn’t tell us why they are that specific way.

Typically, data is a useful tool for optimizing or for understanding what content is performing well or poorly. Based on data, we can understand what/where/by whom/for how long/how often something happened. But it doesn’t tell us precisely why something happened. That’s where we need a human to ask basic questions from another human.

A qualitative study shows us why things are how they are, but we can’t make any generalizations because the sample group is usually so small.

To get the best support for business decisions, we need quantitative and qualitative studies combined with human insight. The more we have and use data to support our decision-making, the more we need human understanding, user or customer interviews, and validations. As strategic leaders know, companies shouldn’t do short-term optimization at the expense of long-term business. Data analytics can be an extremely powerful tool for optimizing business in a short term, but it can’t be the only road sign to follow. The better results we get with data, the more we need critical thinking and insight.

The misleading “happiness wall”

An excellent example underlining the differences between qualitative and quantitative studies comes from the Finnish suburbs. A significant urbanization craze hit Finland in the 1960s and 70s when Soviet-style prefabricated buildings and suburbs were built – even though they had a terrible reputation at the time. Newspapers and general discussions told the same stories of how the prefab-filled suburbs were a cradle of social problems, misery and loneliness.

If we just monitor action, we can miss something very crucial.

Still, all the surveys conducted were showing that people at the suburbs weren’t as unhappy as the common discussion had stated. In fact, the surveys found that people in the suburbs were happy. Matti Kortteinen, a Professor of Urban Sociology at the University of Helsinki, started to wonder why there was such a significant difference between studies and the general attitudes towards suburbs. He went to these notorious suburbs to discuss with residents and collect material for his research. He sat in people’s homes and started to listen. After an hour or two discussions, the suburbanites began to open up. Misery, loneliness and social problems started to arise.

The reason why the surveys had gotten the totally wrong picture was a “happiness wall”. Ask someone how it’s going, and their answer is typically “very well” or “fine” even if the person’s dog just died, their personal and financial life is in crisis, and so on. That’s one way how surveys and data analysis can go wrong.

If we just monitor action, we can miss something very crucial – like Facebook has, and whose purpose is to “give people the power to build community and bring the world closer together” (the exact opposite of what Facebook has done in politics). To understand what are the triggers for a specific action or behavior, we need to study deep feelings, desires, and dreams; we need to have conversations.

By understanding the deeper feelings and wishes that cause action, we can offer something purposeful that builds customer’s loyalty. With a clear and inspiring purpose companies can differentiate from competitors and build long-lasting business.

Time isn’t money

Why Facebook’s case is so compelling is the fact that its popularity is based on a network effect. People want to be there because everyone else is there – but on the other hand, Facebook could collapse just as fast as it came into being – as Wired writes.

However, Facebook has made healthy steps recently, including changing their algorithm to show fewer advertisements and viral videos and more content related to users’ friends and family. But one thing Facebook should still understand is that there is a level when people have spent enough time for one day on the service. Real life and happiness are out there. If we take one hour from someone’s life, we must give something equally valuable back – something with purpose. And that something isn’t funny cat videos.

Time is almost the only thing we don’t get more of in our lives. If we consider how precise we are in how we spend our money, it’s gloomy how little we care about our time. Time isn’t money; it’s much more valuable.

Maija Mäenpää

Designing employee experiences increase commitment and happiness

Maija Mäenpää

What is employee experience and how does it impact recruitment?

Johanna Nieminen

Transformation fitness: Why companies fall into fat diets