Stacey Barr asked:
If you knew in advance how much effort you’d have to put into collecting, collating and managing the data that provides the foundation for your performance measures, you may be put off entirely. Particularly for smaller businesses, departments or teams, you don’t have big corporate business intelligence systems offering you the data you need at the press of a button.
No, you’re often forced to collect and organise the data you need, all by yourself, and on top of your real job! But don’t give up on measuring performance just for this reason. If you’re still a firm believer that meaningful performance measures make your day job better, then these 5 tips for saving time in managing your performance data may just make the difference.
Tip 1: Collect data that is useful, not just interesting.
Market researchers are used to hearing their clients say things like “It would be interesting to also know [blah blah] – can we add a question about this to the survey too?” And good market researchers will say in response, “Not if it’s just interesting. It has to be useful.” They know the time and cost that gets wasted collecting data that doesn’t serve the purpose of the research.
If you collect performance data that strays from the purpose of your performance measures, you’ll collect much more than you need, much more than you’ll use, and probably enough to paralyse your performance measurement process. Be ruthless: collect only the data you know is useful to calculate and analyse the performance measures that really matter.
Tip 2: Build the data collection into work processes.
Imagine you operate a health club and you’re interested in measuring how committed your clients are to their gym programs. The time-wasting way to get the data for such a measure is to devise a survey and get your staff to call as many clients as possible to go through the questionnaire. And this is over and above their everyday work, too!
A time-sensible way is to see where the data might already exist or be easier to get through existing work processes. Personal trainers already fill out gym cards for their clients, to track the sessions their clients complete. Making sure that this data is collected consistently avoids the need for a separate data collection process.
Tip 3: Use a relational database to manage data (not spreadsheets!).
A freight business has their delivery turnaround times not in one spreadsheet, but in dozens of them: one for each month. And they do little better with other performance data such as revenue, delivery misdirections and service costs. They waste hours and hours every month manually organising the data in an attempt to produce trend graphs to report performance.
Even small businesses, departments and teams can recoup the time wasted in manually managing data with an investment in a simple database application, like Microsoft Access. All the data goes into well designed tables, via easy to use data entry forms. Not only is the data all in one place, it’s fast and easy to get access to for both regular performance reporting and ad hoc queries to analyse the data more closely.
Tip 4: Don’t freak out over imperfect data.
A management team in a government agency routinely spends well over half of their decision time debating the quality of the performance measures’ data. What is the real cost to the organisation of losing time in delaying a decision until data is perfect, compared with taking a decision with imperfect data now?
Data will never be 100% accurate, and it doesn’t have to be. Imperfect data can still give you rather reliable feedback about trends in performance. Take a quick check for any systemic data quality problems, to estimate their real impact on the decisions you are taking. After correcting any vital data quality problems, your time is better spent in cause analysis and performance improvement of your business results, not perfecting your business data.
Tip 5: Use samples to estimate, instead of populations to calculate.
To measure the accuracy of their stock management process, an inventory management team were visiting all store locations and counting all stock items held. This took them many months, and almost a full time team to chew through all the counting.
When they sought some advice from a statistician, they learned (with great relief) that they could get a very reliable estimate of their stock management accuracy through stocktaking a sample of stock items at a sample of store locations. In fact, with the smaller task, the counting process had fewer fatigue related errors, and the estimate was not only cheaper and faster, it was more reliable.
Spend your time where it matters most.
Take a little time to think about what data you really need, how much integrity it honestly needs to have to be useful, the easiest and fastest way to collect it, and an accessible way to store it. It will be time well spent, which will create more time down the road for the real work of performance measurement: improving your business performance.