The Wheat from the Chaff #3 - Special

As part of my job, every week I read a lot of content about data, technology, and artificial intelligence (also known as AI). I filter them and send you the best ones, every Sunday.

Hello everyone 👋 ,

This week, we are doing something different. We are releasing an alternative version of “The Wheat from the Chaff” to cheer up your return from the holiday and, of course, keep you informed. Let's talk about an article written by Steven A. Cohen and Matthew W. Granade for the Wall Street Journal in 2018. Despite the date, the content seems like it was written yesterday! It's impressive how it is possible to have a foresight beyond limits.

It is time to separate the wheat from the chaff…

Data-Driven is in the past!

You read that correctly! I truly believe that being Data-Driven is the first step. I often joke that anyone who needs data needs it to make decisions. I know many companies position themselves as Data-Driven, advocating this model as a competitive advantage.

Let's imagine the following scenario: a product (or service) development cycle that is based on statistical models. These models allow for product improvement by capturing more precise data and, consequently, creating new machine learning models that are constantly evolving. This creates a virtuous cycle that continuously enhances the product and, thereby, the customer experience.

It establishes a nearly frictionless process of continuous improvement, as opposed to relying on human evaluations and advancements as in the Data-Driven model. To aid understanding, we can compare machine learning models to a child learning to speak. Just like them, these algorithms learn for the first time and then should be able to generalize what they have learned while also continuously learning something new.

A model-driven company goes beyond a data-driven company. A data-driven company collects and performs statistical analyses to assist its executives in making the best decisions. In a data-driven company, data helps the business. In a model-driven company, models are the business itself.

Currently, much is said about artificial intelligence and big data, but models are the source of power behind these tools. A model, by definition, is a set of mathematical equations or rules that attempt to explain some phenomenon or predict a fact, such as a potential client's default risk or even the next customer. The remarkable beauty of this creation is the fact that it does not require human intervention, eliminating the "human bias" that often makes decisions based on intuition rather than data.

These algorithms are trained to continually optimize results and identify characteristics that no individual could discern from hundreds of thousands or even millions of pieces of information. Once developed, a model can learn from its successes and failures at a speed and sophistication that humans typically cannot keep up with.

A model-driven business then employs these models to drive key decisions in its business processes, creating revenue streams or cost reductions based on increased efficiency. To build this system, a mechanism is needed, often based on software, to collect data, clear processes for model development from that data, and a mechanism to act upon the suggestions of the models themselves. Companies like Netflix, Tencent, and Amazon demonstrate this characteristic, as described by Steven A. Cohen and Matthew W. Granade in their article "Models Will Run the World" in the WSJ.

In the case of Netflix, for example, its recommendation algorithm is famous and responsible for 80% of content consumption. Every time a customer accepts or rejects a recommendation, the Netflix algorithm improves. This transformation is taking shape in various sectors, including some that are traditionally described as traditional: agriculture, logistics, and services.

Who will win this game? Most likely, those who develop the capability to build and integrate statistical models into their businesses, who gather the best data about customers, and maintain themselves within the virtuous circle. In the pursuit of competitive advantage, model-driven companies will accelerate the process in the coming years, and the bet is on model-driven businesses: "Model-Driven."

See you Sunday…

We always arrive in your inbox around 12:08 PM. Some email servers are stubborn and delayed... Others are even worse and throw us into spam and/or promotions. What's the logic behind this timing? ...None :)