Retail Traffic Counter Devices
On average, consumer spend increases by almost 40% when a customer has a positive experience. People counting, or footfall, plays a big role in providing the insight required to create that positive retail customer experience, from expert management of scheduling, to optimizing store performance and assessing the impact of marketing campaigns.
People counting and footfall counters come in a number of different shapes and sizes – here are our top 7, having evaluated many in the market.
1. Thermal people counters
Historically the size and appearance of a small smoke alarm, the thermal camera is now the size of your home router. Fixed to the ceiling above the area to be measured, they are very unobtrusive. They are also very accurate. Thermal counting uses thermal imaging technology to note temperature changes of individuals compared to the environment. Customer body heat is detected by heat sensing cameras, producing reliable data, even where traffic volume is high.
Multiple cameras can be linked to create wider zones, whilst also being adaptable to complex entrances, offering continuous retail traffic counting in complicated environments.
Thermal counters are great where there is a steady flow through the counting area such as shopping malls and, as they do not use visual light, they can work in total darkness.
Ipsos Retail Performance approved/trusted manufacturer: Irisys
Ipsos Retail Performance unapproved manufacturer: None identified to test
2. Stereo people counters
Stereo people counters are typically the size of a rectangular pencil case with two camera lenses. Usually fixed to the ceiling directly above the area to be measured, although the cameras can be tilted to a limited degree and remain accurate. They also have the advantage over thermal devices where the flow through the count area can be more staccato and still the object targets are not lost.
The devices collect entries and exits in real time, collating accurate insight into customer volumes that is adaptable to a wide range of environments. Stereo data is reliable using the same approach as our own eyes to understand depth in a field of view. This allows the device to exclude items by height, for example children or objects such as pushchairs.
An advanced range of stereo tech allows not just entry and exit counting, but also accurate measurement of the number of customer’s and customer dwell time within a defined zone, as well as the ability to path track through the entirety of a customer journey. Only some manufacturers can provide extensive linking of cameras, even in our own suggested Trusted manufacturers.
Ipsos Retail Performance approved & partnered manufacturer: FLIR; Xovis
Ipsos Retail Performance unapproved: Hella, Shoppertrak, BEA
3. Mono people counters
Typically half the size of a stereo device driven by the use of a single lens, as well as being circular and fixed above the measuring area to the ceiling. As it’s not stereo it does not have a depth of view which means children cannot be removed, while the lack of triangulation affects all mono device accuracy.
The range of accuracy at times for these devices can be as low as 50% and, despite laboratory claims, we have not seen accuracy much above 90%. In these higher cases the effort to achieve that is excessive. Prerequisites exist for such devices; traffic is relatively low; lighting is very consistent. The attraction of course with less hardware — one lens versus two — is that it’s cheap to install, however the costs will be hidden.
We have experience in setting up mono cameras and then credibility of the data becomes an ongoing factor which drives a lot of management time for all parties.
Ipsos Retail Performance approved & partnered manufacturer: None have achieved our required accuracy standard
Ipsos Retail Performance unapproved manufacturer: FLIR, Countwise, Axis, FootfallCam, Giken, Modcam, Eurecam, Springboard
4. Time of Flight people counters
Similar in size to the new range of thermal devices and fixed to the ceiling above the area to be measured, a Time of Flight sensor sends a signal out to the objects beneath it and records the reflection of infrared which bounces back to the sensor. This allows a greater depth of vision and movement to be recorded over other devices including stereo and thermal. The device can also operate in total darkness and has the ability to link cameras to cover a wide entrance.
The development of people counting software has the potential to create a richer data-set with not only X and Y dimensions but also Z, which could aid measurement at a shelf level. This is all in the development stage currently and we are waiting for final performance figures to be confirmed and meet our recommended target of 95% lifetime accuracy.
Ipsos Retail Performance approved & partnered manufacturer: Irisys – In final testing and not yet approved for release. It promises to be a huge step forward in both accuracy and ease of installation.
Ipsos Retail Performance unapproved manufacturer: Vadaro
5. Wi-Fi people counters
Wi-Fi operates from Wireless Access Points (WAPs) and these can vary in size, from the size of a cigarette packet to a large home router. They are deployed in the vicinity of the area to be measured, often in the ceiling. The range and accuracy of such systems are varied and require a degree of set up and calibration. Claims of accuracy of sub-five metres are difficult to establish in reality and with the sample set of clients having Wi-Fi switched on this measure is only an indicator and not a true count.
Where this becomes interesting is in the measurement of dwell and patronage, but our advice is that this is a secondary measure of traffic and not a replacement for previously mentioned approaches.
There are many Wi-Fi analytics platforms to choose from so our summary of hardware below comes with a lot more advice and narrative.
Ipsos Retail Performance approved manufacturers:
Meraki: Least hassle, highest reliability. The Meraki AP’s will work out of the box, no additional hardware or software requirements for this. License fees from Meraki include the CMX feed by default and thus no additional license costs. Splash page creation is possible and management for these devices includes social login.
Cisco Aironet: Aironet access points require additional software, which can be hosted externally, as well as a license fee for that software.
Motorola: Supported via Motorola ADSP this could be an additional license fee and can be arranged.
Ruckus: Supported via additional license fees per access point, this can be costly at this stage for large implementations.
UniFi: Cheap access points, fairly reliable and used heavily in South America. These can be supported but require to be flashed with new firmware and thus a site visit!
Open Mesh: Working best as a companion device to Wi-Fi in place, however it can be used straight onto an open mesh network if already in place.
Aruba: Works with ALE using Aruba Hardware. License fees for ALE are normally required.
Ipsos Retail Performance unapproved manufacturer: None identified at this stage but our advice is to refer to our approved list for guidance
6. Infrared (IR) people counters
Infrared people counters can work well but they vary widely in the nature of their effectiveness, largely due to the quality of implementation and the ongoing management which, to remain accurate, tends to be higher than the new generations such as thermal, stereo or Time of Flight, all of which now add remote support capabilities.
Like mono this technology can be very low cost, sometimes even more so, however with all the quality issues the questions remain over whether the data can provide any sensible direction — unless there is comprehensive monitoring and support, something we have developed over a number of years for our clients. In simple terms the approach creates an infrared barrier between sensors, clocking up a count every time this is crossed. While not the most advanced option, IR counters have positive use in a number of limited scenarios; mall Kiosks being a case in point.
Ipsos Retail Performance approved manufacturer: Ipsos Retail Performance ALA; Ipsos Retail Performance ALD; Ipsos Retail Performance Kiosk
Ipsos Retail Performance unapproved manufacturer: a plethora of low cost options which return the same low quality of results
7. CCTV people counters
Utilizing ceiling mounted Closed Circuit technology, often integrated as part of the in-store security infrastructure. This has the advantage of potentially being one less supplier in the mix and often appears to be a low cost item due to it being purely an add on software feature to the already installed security system. That said, the experience of clients we have seen going down this route has been that it’s not as simple as it seems and in effect it suffers in the same way that a mono approach does with lighting variations. In fact, because the cameras have another principle focus i.e. security, the set up always runs the risk of being compromised and as such so does the accuracy.
Usually the set up requires a lot of tweaking for each and every site and between the sites the accuracy we have seen is not consistent. This in our experience has been one of those technology areas that periodically gets looked at for short term savings, only to then find it’s not all it seems and clients then return for a more focused approach with the consistent results required to run their business.
Ipsos Retail Performance approved manufacturer: None
Ipsos Retail Performance unapproved manufacturer: many security based options which focus on people counting and Conversion rate as an additional service
We have stated our opinion on each of the different technologies available. This is derived from spending 25+ years of working with, and testing, most of the available technologies in the market. Others may have their own opinion but for us we wanted to have devices that are accurate, cost effective, easy to set up, remain stable through the life time and allow us to focus on outcomes for our clients rather than spend time discussing individual store/device accuracy which serves little, if no, business value.