Plant Your Data Tree

There’s an old saying that the best time to plant a tree was 30 years ago—the next best time is today. Similarly, the best time to start taking data seriously is long ago—but the next best time is today. If you are in the position of recognizing the inevitability of good data usage as a prerequisite for competitiveness in the future, and yet you know you do not have sound data policies in place today, the best thing you can do is get started on your data journey today. Here are three key steps that we recommend as you get started.

​1. Evaluate and prioritize data points for collection. There is an endless amount of data that you could endeavor to track—a real estate lender might track loans originated, housing units created, risk rating changes, jobs created, per unit development costs, guarantor liquidity, and on and on. To avoid the paralysis that comes with an overly ambitious list, take the time upfront to consider what data would be most impactful for you as an organization, and to prioritize that data for collection. Perhaps you want to focus on data that helps tell your impact story, or data that would help you make better underwriting decisions. By doing this, you will generate a shorter list of critical data with a clear objective—thereby increasing you and your organization’s motivation and ability to carry out the data collection effort. It is absolutely true that by focusing on certain data points today, you may end up having to double back to collect data from the same sources in the future. That outcome is far better than the likely alternative, though, which is to seek to collect all potentially useful data starting now, only to burn out in two months with an incomplete data set when attention gets redirected or motivation wanes.
​
2. Start going backward and forward. To generate a useful data set as quickly as possible, start gathering data from historical sources (old credit memos, reporting that borrowers have submitted in the past, etc.) while also setting up internal protocols for collection of data generated by new events (new approvals, next quarter’s reporting cycle, etc.). By doing this, you will have a ready set of data that can be analyzed quickly, yielding the positive feedback that you and your staff need to sustain motivation and really visualize the benefit that rigorous data collection can bring—actionable data that can help you make better decisions.

3. Implement a periodic review cycle to ensure data completeness. A data set is only as good as its inputs. Missing, wrongly formatted, or inaccurate data can cause errors in calculations, distort results, and degrade confidence in analysis. To avoid this, one person should have responsibility to periodically review a given data set to ensure completeness and data hygiene. The timing of the review should be based on the frequency of data input and the ease with which accurate data can be obtained after the fact. Technical assistance hours, for example, should likely be checked weekly, because if they need to be backfilled any more than that, the person doing so is likely guessing. Information on the characteristics of approved loans, on the other hand, could be checked only quarterly, as the data needed for backfilling are held in credit memos that aren’t going anywhere.

If you execute on this implementation plan, in a matter of months you may have a dataset worth analyzing. Once your colleagues start seeing that you value using the data they are collecting, and you can share with them analytical results from that data, momentum should build further. If you execute on this implementation plan, in a matter of months you may have a dataset worth analyzing. Once your colleagues start seeing that you value using the data they are collecting, and you can share with them analytical results from that data, momentum should build further. From there, the limits of your data usage will only expand in front of you. So, plant your data tree today—there is no better time.