The massive body of data organizations possess is valuable information on their customers, suppliers, operations, and diverse aspects of their businesses. For decision making, the desire to attain the perfect data leads to stalled effort and missed opportunities. Time is spent on gathering as much data as possible, fixing incomplete data, correcting formats, removing unwanted outliers and so on.
The reality is having perfect data is difficult, but there are ways to work with not so perfect data.
Know what you need
What is the problem you are trying to address? When you are clear on that, you would be able to specify the data you need.
For a wholesaler who wanted to determine which customer segments to focus on, they needed to review the revenue growth trends and the corresponding margins associated with each customer segment. Recognizing that pulling data from multiple systems is too onerous if they pursue all customer segments, they chose to focus on customer segments that matter most. This reduced the scope of data extraction and preparation. The additional effort otherwise spent might provide ‘nice to have’ but not critical information for the decision.The additional effort otherwise spent might provide ‘nice to have’ but not critical information for the decision. Click To Tweet
Note the imperfections
Understanding what the data gaps are and their potential effects on the decision you would be making help to flag areas that require adaptation.
Errors in selecting the proper account codes to charge work orders to had been an ongoing issue for a municipality because of the steps involved. The problem affected resource planning for the work crews. Noting that the pending software upgrade would take some time, the operations manager utilized piecemeal data from another source to complement what is available for the main system. The available data are imperfect but the manager was able to plan accordingly and minimized overtime spend.
Make allowance for uncertainty
With imperfect data, the presented information would contain elements of uncertainty. Depending on the level of risk tolerance you have, it would be prudent to make buffer allowance.
A third-party logistics company experienced high growth attributable to the shifting shopping behaviours of the consumers. Capital is needed for warehouse expansion and additional vehicle leases. In evaluating how best to balance their investments, an aggressive projection using the current volume was used to select the new warehouse location. This was due to scarcity of space in the vicinity and the associated costs for relocation are high. On the other hand, the company held off adding new vehicles because it could use contract drivers.
Though it is ideal to have perfect data to guide decision, it is not feasible all the time. Hence, it is important to be able to work with the best, may be imperfect, data available and move forward. Missed opportunities would be minimized.