Machine learning for the masses: Exploration and discovery

By Richard Perret

Nobel Prize-winning author Andre Gide once said, “Man cannot discover new oceans unless he has the courage to lose sight of the shore.”

The typical organization has done much to empower its employees. Most progressive firms allow employees to drive basic decision-making, make financial and purchasing commitments and freely allocate how they spend time.

Yet one area yet to be democratized is access to data and insights. Sure, enterprises provide reporting and dashboards, but those are less about freedom and more about conformity. The advent of machine learning and visualization has unlocked an understanding of the way things work and how to predict how they may work in the future. These are truly tools of discovery.

So, why isn’t there more widespread use of these tools? Why don’t enterprises want to turn more employees into data scientists? Why don’t they want to make machine learning available to the masses?

The typical reasons fall into the three following areas:

  1. Data: There is too much of it, it requires a lot of “cleaning” and we don’t often know what data matters most.
  2. Culture: Organizations lack the courage to empower in a free-form manner.
  3. Skills: There is a lack of analytical skills, let alone data science skills.

So, how can employees discover new oceans — or, if not oceans, new paths to where they need to go?

GroupM is a media investment company charged with making the most of a marketing investment by optimizing the timing and placement of advertisements. Consumer tastes and preferences change dramatically and knowing which factors influence behavior results in improved ad conversion rates.

By empowering their employees, using machine-learning techniques to help wrangle data and predict outcomes, GroupM has seen conversion rates improve by as much as 50 percent. Irrelevant ads are a waste of time for consumers and a waste of money for the brands behind the campaigns. GroupM used machine-learning algorithms to gain greater levels of insight into the targeting, timing and placement of ads to improve their marketing campaigns and ultimately, their conversion rates.

In this video, learn how GroupM found an easy-to-use machine learning tool that impacted business after only one month of experimentation.

This article was originally published on Mobile Business Insights.