Facial recognition: The good and the bad

Facial recognition is heading into retail stores but what does that mean to shoppers’ experiences?

Is facial recognition “creepy or cool?” asked a June 2015 survey of British shoppers. The poll, conducted by product personalisation firm RichRelevance, found that while shoppers were generally OK with personalised recommendations they were a little unnerved by the idea of facial recognition being the source of the targeting. It’s a strange one given that MasterCard recently revealed via CNN that it is trialling the use of selfies as authentication for in-store purchases and Apple has lifted the lid on its biometric authentication mobile commerce software Apple Pay in Europe.

There are natural concerns but while consumers may feel justified in voicing doubt over the idea they are being scanned via CCTV (this is not new – stories about the FBI doing this appeared more than three years ago) they also consistently support personalisation and ease of use when it comes to paying. Something has to give and there is surely a line here which has to be drawn in terms of how data is stored and how individuals are identified.

I spoke to Jeremy Sneller, managing director of TouchByte, a business that develops facial recognition systems for retail.


Do you understand people’s concerns over facial recognition technologies?

I do understand peoples' concern over privacy. Face recognition can be perceived as a scary term as people often automatically assume that those using the technology knows it’s you, all your other personal details and is monitoring everything you are doing 24/7.


Is this not the case?

For the entry level basic analytic data evaluation, this is not the case as the data captured is totally anonymous. All the software is doing is seeing a face in a video stream and counting it. It then estimates the age and gender of that face and stores this data for analytical purposes. The retailer, bank or leisure organisation receives analysis on how many people and in what age and gender brackets visited. This is valuable marketing information but at no point is a person identified.


You said entry level. Is there another level?

Yes, at the next level identification is possible but only when consumers have signed an opt-in agreement to allow their facial image to be used. Examples may be where there is a loyalty scheme offering benefits in exchange for loading your face onto the system. For instance, you may wish your face to be recognised to get swifter and improved service - when you arrive at a hotel, the receptionist is automatically notified of your arrival and your bespoke customer experience is triggered with your favourite drink waiting on the bar for you or perhaps at a car dealership for car servicing, a courtesy car has been brought around for you immediately and the paperwork all processed.


Are there any more levels?

There is a third level which is used generally for security purposes, this ensures only those authorised can enter into certain areas. Again this is to ensure that you as an individual get appropriate protection whether to keep out undesirables or to monitor staff flow.


Is there a danger the technology could be abused or is this in fact protecting people better?

Any technology or indeed any knowledge in the wrong hands can be a potential threat. As a licensed software product, any wrong use could be shut down. The upside is that this technology gives greater benefit to many different businesses, helping make it easier to understand what you as an individual want, make their life experience swifter and more efficient. It can also improve your security, as those people who pose a potential threat can be identified. The technology is there to be useful and as a protection too - as well as to improve business understanding of customer patterns and preferences.


Will facial recognition become the norm in retail now that MasterCard’s interested in the tech?

With a big name like MasterCard showing interest, it will only help the consumers' increasing acceptance of the technology for a wide range of uses. Familiarity and benefit recognition brings acceptance. We believe facial recognition will become more widely accepted due to the greater security it offers, the increased accuracy and speed, the falling entry level costs and the reduced scariness of the term 'face recognition' as people understand the  benefits it offers. People and companies are now more used to having the level of data and information from online systems and it’s a natural progression to have this same level of knowledge from the high street or equivalent.


So what are the main advantages to retailers?

Some key advantages are emerging as this is the era where retailers are expecting more intelligence of their customers both online and in-store. Some examples of where there have been considerable advantages are greater knowledge of customer footfall by a period, increased accuracy in the footfall figures as staff, security and regular suppliers are enrolled on the system and excluded from the figures, monitoring the success of any national or local press advertising in the footfall generated, the ability to monitor and validate opening hours which might vary depending on footfall by each store over a week or month, determining average customer dwell time and movement, understanding age and gender profiles, staff demographics can be matched at different times of the day to help maximise sales as people often prefer similar or more appropriate staff who can relate to their needs or tastes, plus of course an understanding of the impact of external influences such as seasonality and weather.


What about other sectors or is this just a retail thing?

Banks with branches are similar to the retail environment so they will benefit. Leisure organisations – hotels, pubs clubs casinos and stadiums - all need more information about their visiting customers. For hotels, we are looking at a fast check-in process. By uploading your face picture to the system prior to your visit, on arrival you could immediately gain access to your room, hotel facilities, the gym, restaurant etc.  Also face recognition could help hospitals ensure their patients are safe, in the same way educational establishments can ensure the safety of their students. So there are many varied opportunities to develop solutions within a variety of sectors.


It’s not a cheap way to analyse your customers, I guess, so is there a good return on investment period?

The investment payback is all around the interpretation of the data captured and presented and its intelligent use. Also can businesses afford not to know the type and number of their customers? Increased sales result when businesses understand and can plan for the number of customers expected at any point in time; they can ensure staffing levels are adjusted accordingly to maximise the sales conversion into revenue. This can easily be historically tracked and a forecast made on trends. Ensuring the right demographic staff member matches that of the customer base helps put the customer at ease and maximises the customer spend. Average spend per customer can be tracked and incentives introduced to encourage a higher conversion ratio.


So what is the biggest challenge? Technical or cultural?

The biggest challenge is not technical as the technology has been developed and used for national security and border control for many years. It is the acceptance of the technology by the various demographics ages that is perhaps the biggest challenge, such as the 35+ consumers. The under-35 consumers are confident with sharing levels of data, such as posting “selfies” on social media, so they have no issue with using their face for analytics and recognition. The 35-50 age group is less at ease with this and more sceptical whilst in general those over 50 are more resistant.


Does this have potential to feed into smart cities and the Internet of Things? If so, what is the benefit?

Integration into smart cities is already being considered. By identifying individuals as they arrive home, the technology can recognise them and trigger various actions – for instance opening the door, switching on the heating, turning on the TV, boiling the kettle etc.  It allows secure access to currency at the Bank ATM, it could open your car, set your radio station, seat and destination. By incorporating the technology into the city infrastructure it also can provide a safe and secure environment for people to move around and feel safe in. Intelligent street lights are being considered as a way of both monitoring footfall in and around the city and also for use in recognising undesirables and trouble makers.