Garbage In, Gold Out in 3 Steps

by Fred Shilmover, Founder & CEO, InsightSquared, a Bullhorn Marketplace Partner The worst enemy of good analysis is bad data. As the old adage goes, garbage in equals garbage out. No matter how diligent the members of your team (sales, recruiting, finance, etc.) are, given enough time, data entry errors always creep in, which means your reports will become increasingly inaccurate, and you will begin to make bad decisions about your business. But this does not have to be the case. Outlined below are three concrete steps you can take as a manager or owner to improve the data that drives your ability to make good decisions. Step 1. Pick Your Battles You have a ton of data, but most of it doesn’t matter for tracking the key performance indicators (KPIs) that drive your business. Don’t ask your team to spend time meticulously maintaining information that doesn’t ultimately and tangibly drive performance. To pick your battles with data entry (which is what you’re really fighting), start with this quick checklist: • Identify the important values and activities • Don’t ask your team to do any more than the minimum • Start with KPIs that have a quick and visible payback for the team Step 2. Make Data Quality a KPI If you plan to compete on analytics, then data quality is a performance driver. In a weekly review with your sales rep you might say: “Hey John, great job placing those 2 candidates last week, but I noticed you have 3 data errors that are impacting your performance metrics this week already. I need you to clean those up today.” Use errors in the indicators you selected from point 1 above as a negative when evaluating performance. Among our clients, we tend to see that the majority of data errors are relatively quick fixes: a mis-typed salary range on a job, a missing end date on a job, a misplaced decimal on a fee arrangement, etc. The typical staffing firm has data errors on 60% of their job orders. (That is a lot of errors!) Step 3. Fix It With Smart Guesses If your average placement is $20k and you have a single Job Order in your pipeline worth $2M, you know that something is wrong. Or if a pay rate for a contract job is blank, you know that the job isn’t going to pay at 100% margin. So what can you do about it? Fix this with a good guess. For example, take your average margin and the value of the Job Order, or if you’ve done enough business with that client, you can take their average margin or if not, the average margin for your collective client base. Think you can do this in Excel? Probably, assuming that you are good with Excel and have the time. Could you write a few rules down that you could delegate to an admin to do every week? Probably, as long as there aren’t any anomalies. Think someone could write software to do this in an automated fashion, with better precision, available on demand? We do too. Imperfect data should not prevent you from being analytical. Use this 3-step approach to get some quick wins and get on the analytics band wagon.