Blake Laufer

 Unless you’ve been living under a rock for the past few years, you have undoubtedly heard the term “Big Data” thrown around. But what is it exactly? And does it live up to the hype?

Today’s modern parking manager is being hassled by upper management — a director, a mayor, or some other bigwig — to bring forward hard facts for decision-making. It’s no longer enough to have experience and an opinion when making changes to your parking operation. These must be backed up by charts and graphs and the data that generate them.

The first challenge with so-called Big Data is that there is no meaningful definition. It’s a loose concept, and one of the latest being thrown around in executive offices. It’s a term you hear frequently, but likely don’t have a framework to apply it to your work or operation. 

Big Data describes the mountains of data collected by your technology and systems. But how much data make it “big?” More than what you are accumulating today, if you listen to the experts. Additionally, what that definition implies — but does not clearly explain — is exactly how your data can deliver true value to you and your customers. 

Data are not information. Raw data are dumb; information is knowledge. Therein lies the promise and pitfall of Big Data. Simply having mountains of data is not enough; you also need the right tools to mine it to find valuable gold nuggets.

How we make Big Data productive depends on our ability to turn data into information. 

If you’re looking for a needle in a haystack, then you want the right tools — perhaps a strong magnet. If you’re looking for gold in a pile of ore, you might consider a metal detector.

Business intelligence is one such tool for the Big Data haystack. BI, as it is more commonly called, is able to take large sets of unstructured data and turn them into information — actual knowledge and insights.

Parking operations are familiar with so-called BI tools, though perhaps not under that name. BI tools include software and systems that produce dashboards, forecasting and prediction about your parking operations. They help you see the patterns or relationships in your data that can lead to action. BI tools do more than reporting — reporting tends to look backward while BI looks forward.

By understanding the relationships in your data, you can take action to achieve results.

These are the types of data that become truly valuable for decision-making, and subsequently convince the parking director or mayor that a certain decision or policy is right or wrong.

Here’s a simple example: Let’s say you’re trying to gauge the success of your parking enforcement officers (PEOs). You ask, “Who is my best parking enforcement officer?” 

The first thing to determine is how best to measure success of a PEO. In some cases, you might ask which parking officer issued the most citations; this isn’t a fair measure, because some officers spend more time on enforcement than others. But if you divide the citations into issuance per hour, now you’ve got a more level playing field.

Citations-issued-per-hour is a great starting point. The “heat map” chart here shows some results. The citations/hour calculation is simple: the total number of citations issued in a shift, divided by the time of the shift. The darker the cell (heat), the higher the number of citations.

By examining these data, it’s easy to see that Officer J is a consistent performer, while Officer T has some highs and lows.

What makes this actionable is the opportunity to look at the poor performers (A, S) and ask why this is the case. Using data as your guide, you could ask officers A and S to swap enforcement routes with J or T and watch how the citations/hour changes over a few months. Additionally, you could examine how many citations are appealed; maybe the high performers also write sloppier tickets, leading to higher appeals and more downstream work for the parking office.

In addition to looking at past performance, this chart can be a predictor of the future. This image shows the same 11-month time period (values omitted). By slicing the numbers by time of day, a similar analysis suggests more enforcement may be needed at noon and 3 p.m. when fewer citations are written, but there is plenty of saturation at 10 a.m. An actionable item could be to stagger PEO shifts, so there is more balanced coverage over the course of the day. Furthermore, far fewer citations are issued on Friday; more investigation is needed to gain insight and determine what action is appropriate to test or validate this data.

You’re collecting lots of data on your parking enforcement efforts — officers, patrol routes, citations issued, and appeals. When you look at the full picture, you can see how Big Data through a business intelligence lens can illuminate actionable solutions. 

In addition to business intelligence, benchmarking is another useful data mining tool. In parking, we’ve seen many formal and informal benchmarking initiatives, and until now, these have not been widely successful. The primary difficulty is the absence of standard measures, and the secondary difficulty is the variability of parking operations. So far, across the parking industry, we’ve been comparing apples and oranges.

One advantage of Big Data is that when a sufficient number of parking operations begin to track their data the same way — often using the same management systems — it becomes possible to compare apples to apples. As an industry, we’re not sufficiently advanced to compare a Granny Smith to a Golden Delicious, but with data-driven decision-making tools, we’re getting closer to recognizing that they are apples.

As an example of simple comparison, this chart suggests parking density across eight different parking operations. Two simple metrics are used for this calculation: the number of full-time equivalent staff in the parking operation and the total number of spaces under management. Calculating the ratio of spaces per FTE is simply a matter of division.

Interestingly, this division shows two clusters of data. There is a group of three higher-value ratios around 384 spaces per person in the operation, and there is another group of five lower ratios in the neighborhood of 161 spaces per person. Would it surprise you to learn that the three are downtown, urban operations while the other five are suburban parking operations?

Benchmarking in this manner gives you the data needed to make a case to the director or the mayor. Armed with the data here, if your urban parking operation runs at more than 400 spaces/employee, then you might have a case to hire another staff member. This is why it’s so important to compare apples to apples in parking segments and across the industry.

In conclusion, Big Data is not the answer. Business intelligence and benchmarking provide much more insight into your parking operation, and each requires far less data than one might expect. Actionable data already exist in your parking operation, and we don’t need to call them Big Data for them to be useful.