Data-driven tips: accuracy

Number four of the data-driven tips for 2023 is accuracy. In this blog, Data Strategy Leader Louis de Roo tells you more about how the accuracy of your data ultimately helps you make better decisions.

The more accurate your data, the better your forecast, right?

Anyone starting to work data-driven often faces the challenge of switching from decisions based on gut feeling to decisions driven by data. And that's where the discussion begins: because when is data good enough to base decisions on? The more accurate your data, the better your forecast, right?

Not necessarily. Because the only truly accurate data is data after the fact. That's of course quite useful for evaluating your decisions afterward and learning for the future, but you'd rather base your decisions on data that actually looks ahead.

Because the future is hard to predict, forecasts about that future are, by definition, actually inaccurate. And that's where many managers get a bit nervous; because what if the forecast turns out to be wrong? Or could have been better? What if, in hindsight, that decision turns out to have been the wrong one?

Data as a navigation aid

Let's first take a deep breath, and realize that a decision based on gut feeling is still less well-founded than a decision based on inaccurate data. And that it's — of course — wise to align the impact of your decision with the accuracy of your data. If your data isn't all that reliable, be a bit cautious about making very drastic decisions for the future of your organization. See data about the future as a navigation aid, not as a conclusive forecast.

KPIs

We previously wrote about the value of the right KPIs to help you navigate. But how do we get away from the constant tendency to view KPIs only as an after-the-fact assessment?

Anyone who wants to steer their organization with data will need to look for the KPIs that enable decisions. And to find those, it's important to know your business processes, and to understand where in those processes the essential decisions take place. The annoying thing is that most business processes tend to quickly become quite complex, and that especially the number of exceptions to the basic process quickly becomes unmanageable. Searching for the right decision points in your process, you quickly lose sight of the forest for the trees, and it seems easier to just look at the process afterward and assess it with a "lagging" KPI.

It's precisely that complexity that makes the added value of data-driven decision-making so significant. Because by basing small decisions on data at the right points, you carry out dozens or even hundreds of small process optimizations. Which, of course, translates into increased efficiency, more profit, or higher customer satisfaction.

Process mining

Wonderful visions of the future, but fortunately far from unattainable. Where mapping out a process in detail used to be a weeks-long job for a business analyst, with modern process mining software it's now a matter of a few minutes to get a clear and organized picture of your process and all its deviations. In fact, with that analysis in hand, it's a piece of cake to estimate the impact of your process decisions. Do your orders keep getting stuck for days on a manual transport price? Or do you pay some invoices twice because they're booked under two categories? Making irregularities in your process visible is the first step toward improving decisions based on data — and toward optimizing your process!

Knowing your process steps and their quirks enables you to figure out within which bandwidth certain figures should fall for the process to function optimally. And at which deviations you'd want to reconsider the decision.

Mapping out those deviations is your first set of "leading" KPIs: inaccurate, but closer to the decision. Which allows you afterward, with your "lagging" KPIs, to accurately determine that you made the best possible decision.

Written by Louis de Roo
Data Strategy Leader
E-mergo