The efficacy of critical decisions is based on the reliability of the measurements and supporting data that informed those decisions. Data reliability is therefore crucial. However, there is a surprising level of confusion related to how we characterize, assess and even manage our data.
One of the aspects of data curation that is frequently misunderstood – especially in the financial disciplines – is data precision. Precision is helpful, most certainly, but not appropriate in all situations. Precision expresses the ability to define a measure or a characterization in extremely fine detail. However, hyper-fine detail is usually unnecessary to evaluate trends, understand states, recognize patterns, or draw informed conclusions. Additionally, precision almost always requires higher investment in both time and/or tangible resources before the data is “ready for consumption” making it expensive and likely to incur delays.
The last notable danger of precision data is that it must be additionally screened for errors. We cannot afford to allow corrupt data to coexist with critical data but it is usually the case. How clean is your database, your measuring systems, or your data collection processes? If the data has trash or errors in it, no amount of decimals will improve the insight it provides. If AI is something you’re toying with, this is particularly critical. Be warned.
Interestingly, precision is generally an accepted form of waste in a decision-making process if we recognize that there is a more efficient, effective, and timely alternative. As waste it is ripe for elimination in order to improve the performance and output of the process. Any form of waste consumes resources and delays outputs. What then is the more appropriate alternative to the notion of precision?
We’ve asserted that precision represents the most detailed approach to data management. Accuracy, on the other hand, focuses on the trueness, validity, and quality of the data. We use the term accuracy to assess for purity of the data and to get an objective sense of how relevant the data is. Timeliness is one of the helpful features of accuracy that makes it so well-suited for critical decision-making.
An analogy that sticks with me is that of a target. Precision is the tight grouping you would like to achieve but without much respect to the bullseye. Accuracy, however, emphasizes our ability to get to the true objective – the center of the target – rather than just having a tight grouping. How true and “on-target” our measures are will ultimately inform our decisions better than the extent of the detail.
When important decisions are being made, are your inputs timely enough? Are they hyper-precise but not entirely reliable? Are your measures suspicious or don’t always seem to adequately explain what your business is experiencing? You may have inadvertently substituted precision for accuracy. Consider deliberately moving toward real-time visual measures that help convey accuracy without the pitfalls of excessive precision.
Lean in and Lean on.