Data-driven decision making is highly valued. After all, if hard facts are the basis for improvement, success will surely follow. However, the insights you are relying upon can only be accurate if the data input they are derived from is properly understood, in context. In my experience, many organisations fail to get the results they expect from data-driven decisions because their interpretation of the relevant data is flawed.
Why does this happen? Mismatches in descriptions, definitions and attribution can turn certainty into chaos. The good news is that these things can easily be fixed if you take just a little time to examine how your data is defined. I’ve put together three simple checks which you can do in just one day, so you can get straight down to taking corrective action!
Check 1) What’s in a name? Clarify your report descriptions
When comparing reports used within your company, do they all use a unified language? I have often seen different names for the same thing depending on who created the report, the system used, or the time when the report was created.
For example: terms such as vendors, suppliers and accounts are often interchanged even though they all refer to the same thing. Or one process has a start time, while another has a begin time. This inconsistency is confusing to all those who work with your reports, and can lead to report findings being incorrectly interpreted.
What to do: identify where multiple terms are being used for the same thing and define which is to be used in each case. This will unify your report descriptions so everyone has the same understanding of what they mean.
Check 2) What’s your definition? Show calculations and sources
When you start asking report creators about the details of their content, you get some very insightful answers. Because while a number of those involved will be familiar with how each definition is derived, most of the people reading the report will not. This leaves far too much room for interpretation, with users not necessarily attributing the data to the correct context.
For example: when the “number of orders” per period is stated, does this refer to the number of orders placed by a customer, or the number of orders produced, shipped or delivered? Does the figure refer to every order that existed, including those which were scrapped, or just those which went to completion?
What to do: ensure that insight into how report content has been generated is available to all users. Add this information to your reporting tool, so that it is available to all users at all times.
Check 3) Is your administration aligned with your process? Improve the timing.
So, you’ve unified your descriptions and you’ve clarified your definitions. The last thing to check is whether the timing of data actually corresponds to the physical processes it relates to. I’ve seen plenty of examples where this isn’t the case, which leads to incorrect understanding of workflow and stock control.
For example: delivery to a customer is not logged until the courier returns to the depot. Or receipt of raw materials is not booked-in until peak inbound workload has subsided.
What to do: ensure your information logging processes are aligned with your physical processes, so data related to activity timing gives an accurate picture.
Investing just a single day into addressing these three checks will significantly improve your organisation’s ability to understand its data correctly and make watertight data-driven decisions. I’d say, give it a try!
Chris Beckers is Business Process Consultant at R&G Global Consultants in The Netherlands