Understanding how your customers use the services and facilities that you provide for them provides you with invaluable insights into how and why they might be using them. You could be sitting on valuable data that can be used to help you understand more about how your customers use your products or services. There may be clues about additional services that you could offer and the good news is that the cost of acquiring these insights has been falling and will continue to fall.
Quite recently, I was approached by a large shopping centre operator who has been receiving complaints from customers saying that there a lack of available parking in one end of the complex. This end has had quite a few new tenants move into the complex as well as a large hardware store start operating. The mall owners were of the view that the hardware store was the root cause for the parking issue but wanted to use data to back up their decision-making.
They wanted, like most businesses to get a better understanding of the people using their carpark. Firstly, they wanted to understand how many unique cars were using the car park, the daily volume of cars, how many cars were appearing on multiple days and were there people coming to the centre multiple times within a day. Secondly, they wanted to know who was using the carpark, was it mainly the customers of the centre, or was it the employees of shops in the centre that were taking away parking spots from customers. So, they wanted to see if there were recurring parking patterns. Lastly, they wanted to understand was how long people were staying at the centre.
Fortunately for the centre, they could capture this data quite easily by installing cameras at the entrance and exit points for the car park. These days cameras designed for carparks come with the ability to not only capture images of the license plate of a vehicle but also, they have License Plate Recognition (LPR) that allows that image to be turned into text along with a timestamp of when that event happened. The centre had the cameras operational for less than two months and during that time period they had generated just under 100,000 records which provided some challenges for the analysts at the shopping centre but was a medium sized data set for us.
The technology to capture customer data points has improved massively in the past few decades, the cost to capture that information has fallen significantly and most devices which have sensors have the capability to export the data they generate so that it can be analysed in other applications. Given the ongoing reduction in cost of capturing customer data, it may be worth considering how you could capture useful data about how your customers use your products or services that can provide insights for you. The first business in your industry that unlocks these insights and uses them will have an advantage over the rest of the industry. It would be great if the business that does that is your business.
We needed to do some transformations to the data set provided, e.g. we needed to extract all the unique license plates being captured, set up views of the data by date and day of week and create the data pairs – carpark entry and exit times. We needed to write some custom code as we found that some cars were using the carpark multiple times in a day and that the cameras were not always capturing both the entry and exit events as not all cameras were working for the entire observation period. So sometimes we had odd numbers of events related to a vehicle so we knew a data point was missing but we weren’t able to tell whereabouts in the day that had happened.
Whilst that meant that not all data could be used, most of it could be as most carpark users (>99%) only used the carpark once in a day and we filtered the dataset so that we only used data pairs, i.e. there was both an entry and an exit of that vehicle.
And, what did the shopping centre learn? Well, they found out that whilst there were some shop employees using the car park they represented well less than 1% of the carpark users. They found out that there were more customers visitors to the car park than they had originally thought. The gaps in the camera data weren’t identified which led to them initially low-balling the number of visitors. They learnt that there some distinct segments within their customer base – frequent visitors, recurring visitors and occasional visitors and that the length of stay was inversely proportional to the number of visits to the centre. And though frequent visitors did tend to stay longer than the other cohorts they didn’t stay that much longer.
What would you want to know about your customers? Have you thought about what type of sensors (cameras, motion detectors, heat sensors, moisture sensors etc.) could capture the data you need that would allow you to understand your customers better? As mentioned, sensor capabilities are increasing and the price of each sensor is dropping so at some point in time in the not too distant future it will become economic to start capturing this information about customers in your industry. We think that it’s worth keeping a tab on this, as you certainly don’t want to be left behind if your competitors start to make a move to tap into this knowledge before you do.