Decision-Driven Data Analytics: A Gourmet Meal or Leftovers?

When considering automated decisioning systems, the starting point should not be data analytics. Begin by asking what routine, automated decisions your business has to make – and only then move on to the question of data.

When you want to cook a wonderful meal for someone special, you don’t just open the fridge and stare at the half-empty shelves thinking: what can I make with this? Instead, you ask yourself what will they love and what else do I need to make it?

Yet when thinking about data analytics, many businesses choose the leftovers option. They look at the data they already have and think: what can I make of this and what can I learn from it? But it is far more effective to think about the urgent or routine business decisions that have to be made, and then determine what data you need (or already have or need to elicit) in order to make them.

Looking more closely, it’s clear that the leftovers approach doesn’t work in any industry. The tabloid press don’t simply ask: “What happened yesterday and what can we make of it?” Instead, they ask: “What do people want to read and where can we get the ‘facts’?” Tesla isn’t successful because it already had a huge assembly platform and then asked: “what kind of car can we make with this?”. Instead, they built the car that they thought would excite people – and sourced or developed the components they needed. The iPhone wasn’t an evolution of existing mobile phone technology (“what can we make with this?”). Instead, they wondered how to make the ultimate mobile device that would excite consumers, and then procured or developed the technology they needed to produce it.

Decision-driven not data-driven

Even when making AI-powered, data-driven and automated decisions, success depends on how you approach these aids. If you only look at your data and rely solely on data analytics – so your approach is 100% data-driven – you will find it hard to make any great leaps in your product development. Indeed, people often try to use the available data to explain and justify their decisions retrospectively. But at the end of the day, this is throwing money down the drain. It does not lead to disruption or true digital transformation.

So if you are looking to enhance your business by introducing automated decision-making systems, it is best not to focus too hard on the data and analytics you already have. Instead, think about the decisions you need to make before considering what data you have available or will need for this purpose.

The table below illustrates the differences in approach:

Decision-making rather than pure data analytics

This approach is particularly useful when hundreds or thousands of decisions have to be made automatically every day, ideally with no need for human intervention (because this runs counter to scaling and automation). Analyzing data (not an end in itself) and learning from it is not the number-one priority. The key is to make decisions that generate business. Making these decisions requires a system to be fed with appropriate, valid and meaningful data.

Therefore, when starting out with a digital decisioning platform, the first priority is to ask: what decisions have to be made? Only after that should you ask: what data is needed to do this? In this way, digital decisioning platforms are the ideal drivers for designing, managing and automating complex operational decisions.