Rewatch COVID-19: The New Normal Webinar. More Information

The Differences Between Manual and Automated People Counting

In this age of automation, I still find it surprising to see people with a manual tally counter or a piece of paper, trying to capture the number of people entering and leaving a premises. I have been exposed to automated counting for over 12 years now, but I do...
5 min. read
Share on
Share on

Listen to the article

In this age of automation, I still find it surprising to see people with a manual tally counter or a piece of paper, trying to capture the number of people entering and leaving a premises.
I have been exposed to automated counting for over 12 years now, but I do remember how surprised I was when I interviewed at SMS, to see that devices to count actual people existed not web visitors (which I understood quite well). It was most likely way more surprising 12 years ago than it is now but I would dare say, judging by the reaction of people when I explain what I do, that there is little knowledge that automated people counters exist among the general population. Sure, it is way easier now to just search for it and see if it exists. After all, our brains are now wired to think that if it concerns collecting data, then someone must be doing it. After all, we live in the age of Big Data. The most surprising fact is to realize that retailers have been doing it for over 30 years, but that outside of mid and large retailers, it is still not yet common and/or fully understood.
Hence, I always smile when I see someone standing, with a small tally counter, clicking away every time they see people coming in a door. You can see their eyes scanning the group of people, and you can sometimes spot the desperation; Did I count this right? That usually tells me the guy is new to it, because the ones who have been doing this for a while just clicks away, with no emotion on their face, and their eyes wandering around, looking at everything but the people. Either they have become experts, or they are bored to death and realize that it does not matter if they are off, after all, who else is going to sit there and count the group again to prove them wrong?
The other scenario I saw in a furniture store, was that the receptionist was responsible for capturing the number of people entering on a sheet of paper. As I was talking to her while getting ready to pay for my transaction (nice home theater seats), I saw her notice 2 persons coming in, which she quickly wrote down. But then, a minute later, while she was processing my credit card, she had to turn around to enter the information on the handheld credit card machine. During that exact time, 2 more people came in to the store. I waited to see if she would realize it, or notice them later, which she did not.
You are probably getting the feeling of where I am going with this. I decided to write this blog for 2 reasons. Let’s start with the first: once again today we got an email saying this from one of our new retail customers:
“Our VP of sales is doing a manual audit in several stores between 9am and 3pm and comparing the numbers with the dashboard. This would be a major black eye to the project if those numbers don’t match.”
That usually never works out. And in the end, the people counter, the “machine”, in this case, the 3DScope II LC, comes out on top over the human, let me explain why:
1. The machine does not get bored, tired, annoyed. The machine does not get sleepy or start to daydream because it partied too hard yesterday evening or had to take care of a sick child.
2. The machine does not make judgment calls outside of its programming. Did I count those 2 people I see over there? I blanked out for a little while, I do not remember. Ok I’ll add 2.
3. The machine does not need to take a bathroom break.
4. The machine does not need a break to eat.
5. The machine does not start talking with customers or coworkers. Let’s hope that never happens by the way.
6. The machine does not have a bias. This is not always the case, but a story such as this is not uncommon either: The store manager told me to count less, so that if the machine is discredited, she will have an excuse against the constant pressure on his “low conversion rate”.
7. The machine may have features such as height filtering to determine if it’s a child or not. Or a feature to filter out U-turns (customer coming in but then leaving automatically). Whether the camera records a count or not will depend on how it is set up. Hard for a human to match this.
I hope I made my case and that you laughed at one or two of the above. On a serious note though, you should expect an automated people counter to count accurately, hence why validating it is a good thing. So how do you do this? You leverage the feature of the people counter, assuming it’s a 3DScope II LC (yes, yes some self-publicity here), and schedule the recording of a video (yes you can do that with a 3D Scope II LC!). The video includes the count lines and the count number, we can provide a special software tool making the validation easy. Record up to an hour, although 15 to 30 minutes in peak time is usually sufficient to build trust. Then validate the video while making sure you understand how the camera works (we can help with that). Then compare the results of the live video with what the camera has recorded. That way you can rewind, when you are unsure and recount. If there is any problem, then usually it’s a minor adjustment which we can help you with. The majority of the time, it will have counted as it should. The good thing is that you will have a video proof to prove it. That usually helps in removing the doubt!
The second reason is simple.
Do not count manually. Devices to automate counting are now more affordable than ever. Paying someone to spend a part, or all their time, trying to capture how many people came in, will not produce a reliable result and will be a waste of money.
Oh and about being affordable… and automated. Stick around as we will announce something quite revolutionary in the coming months!

Share on