• Tuesday , 21 November 2017

Retail signage: it’s all in the numbers

Remember when retailers used to justify their investment in digital signage with the savings on printed posters? It’s still a valid argument, but these days. Customers want a bit more for their money. To supplement the data flow from RFID tags and loyalty schemes, retailers are looking to action a bidirectional flow of information about their customers from their digital signage installation – and, at last, it looks like that’s becoming a reality.

audience measurement
There’s only one problem with being the first in a new field, there’s no one else to share the burden of creating a new market or to educate potential users. For a long time now, NEC Display Solutions has been something of a voice in the wilderness. Its Leaf Engine technology, is a middleware that integrates more than thirty different sensors in order to provide context aware signage experiences for customers, has been the centre of many tradeshow exhibits, and has even been featured in the company’s Ruislip foyer, but has not seen a great rush to adoption.
The logic behind the Leaf Engine solution is indisputable. As Thomas Walter, Section Manager, Strategic Product Marketing, at NEC Display Solutions Europe, explains: “Instead of simply delivering standard content, it is tailor-made and directly addresses the needs of the audience. Through interaction, the engagement of the audience increases, viewing times are extended, buying rates see significant uplift and the investment into the displays quickly pays off.”
Deployment
Now, there are even case studies that illustrate use of the solution and prove the concept. Walter continues: “This is perfectly demonstrated by the Clas Ohlson concept shop, where NEC partnered with SMS/Smart Media Solutions and Smartsign to create an interactive digitised display solution. The potential customer is enticed to try the window cleaning product. By engaging with the product. he is convinced by its benefits and makes a well-informed buying decision.” (The installation can be seen in actions at https://www.youtube.com/watch?v=RRXDvE6rEIM).
Walter also believes that the concept perfectly matches the idea of the Internet of Things, whereby items become part of the transaction and become integrated into the signage concept. But, a category of one, however innovative, is a pretty lonely place.
Increasing the field
Then last month, AV News reported that Signagelive and AdMobilize had formed a partnership to integrate audience analytics with the newly-announced Signagelive Marketplace. The importance of this announcement is twofold: first, in order to supply tailor-made content the signage operator has to know a lot about the viewer (hence the need for analysis); and, second, Signagelive has a substantial installed base, so increasing the potential for the adoption of the technology.

Interaktiver-Screen-bei-Clas-Ohlson-
 Clas Ohlson install
AdMobilize describes itself as a global leader in computer vision intelligence, dedicated to making sense of the physical world, to bring real-time audience analytics and dynamic creative capabilities to digital signage networks. Further, AdMobilize is a transformative analytics platform provides data to understand attention, demographics and emotions of viewers engaging with networks powered by Signagelive.
AdMobilize says it provides the DOOH, OOH and retail industries with the most complete and turnkey solutions including: audience measurement and computer vision solutions, including face detection, vehicle recognition and crowd analytics. The company equates its highly scalable solutions to “Google analytics” for the physical world.
Jason Cremins, Founder/CEO of Signagelive, explains: “Being able to support the proof of play and proof of display data that we capture in Signagelive with proof of view, is imperative to Signagelive resellers and network operates who want to measure the true impact of their digital signage. AdMobilize provide a simple, powerful and cost-effective audience measurement and visual analytics solution that both compliments Signagelive, and meets the demands of our partners and the 1,800 digital signage networks powered by our platform.”
Mike Neel, Global Head of Marketing/Sales at AdMobilize adds: Working with Signagelive and their network of partners just makes sense. The demand for audience measurement and visual analytics for digital signage has reached a tipping point, and our technology allows it to be highly scalable. We believe our combined effort will make utilising these solutions more practical and effective for networks. The ultimate goal is to provide insights that positively impact their digital signage solution.”
How it works?
The AdMobilize technology detect faces within a frame, generates key points on the face and tracks dwell time and session time. Data derived from detection is passed to ‘neural networks’ to generate demographics using a hierarchy of features (age, gender, and emotion for each detected person), to compare to an existing model of what the demographics should look like. These metrics are pushed to a dashboard to trigger web hooks.
Having been a victim of many misclassifications, the question is: “Just how accurate are these machine-generated statistics?” According to AdMobilize, pretty accurate, with scores for emotions ranking 80.5% for ‘happy’, 83.3% for ‘angry’ and 91.1% for ‘surprised’. Interestingly, the most accurate metric was for ‘disgust’ at 98.5% (although we are guessing that we probably wouldn’t need to wait for our digital signage to tell us that one.)
Differences in mood can be subtle, but it is a failure to identify such basic indicators as sex and age that usually prompts (a) hilarity and (b) offence – neither particularly desirable when displaying content designed to exert a positive influence on a customer. AdMobilize claims a 91.3% accuracy for both men and women and says that this, along with all the demographics they detect is improving all the time. Age detection is accurate to plus or minus 10 years.
Tipping point?
So, is Neel correct when he says that the market for signage-based audience measurement and analysis has reached a tipping point? Signagelive and AdMobilize have several case studies in the works in partnership with VARs, MSPs and Networks for the remainder of 2017. Both companies look to grow the number of projects throughout the year and continue to provide a more seamless solution that continues to evolve the convergence of data and digital signage.
With the increasing accuracy of audience measurement and analysis, and the improving performance of the technology, it is fair to say that retailers are increasingly willing to trust the data that the solutions are generating. The result will be better decision-making in marketing, stock selection and store layout. Rather than: “Do we need to integrate audience measurement and analysis into our digital solution?”, the question is now moving to: “Can we afford not to integrate audience measurement and analysis into our digital solution?”

admobilize-500x300“The demand for audience measurement and visual analytics for digital signage has reached a tipping point, and our technology allows it to be highly scalable.” Mike Neel, Global Head of Marketing/Sales at AdMobilize.

Cash-only retailers lose business

The impact of digital technologies on retail was confirmed by news that the traditional model of cash being exchanged for goods at the till could actually be turning customers away from bricks and mortar stores.
An independent survey of 2,000 UK adults commissioned by Ubamarket revealed that 73% of shoppers across the nation – the equivalent of 37.5 million people – have changed their mind, and decided not to buy something after seeing the size of a shop’s queue. The nationally representative study by the retail app also revealed that ease of payment has become a make-or-break factor for retailers seeking to gain and maintain shoppers.
Ubamarket’s research found that retailers risk frustrating would-be customers by maintaining archaic payment methods – over 10 million of us (20%) will consciously avoid shops, restaurants, newsagents, cafés or bars that only accept cash. Moreover, 32% of Brits actually select online or high-street retailers based on how easy it is to pay for items.
The nation’s frustration towards outdated in-store payments also extends to self-scan machines, with 48% of UK shoppers frustrated by the current iteration of self-payment infrastructure in Britain’s retail outlets.
In app payment is found to be a timely solution; 41% of us – the equivalent of 21.1 million people – expect that all payments will be processed via mobile devices in the future. According to the research, in app payment is already the preferred method for one fifth of all shoppers, rising to over a third of millennials.
Will Broome, CEO of Ubamarket, commented on the findings: “Over the past decade, major retailers have taken some strides to improve in-store shopping by embracing new technologies. However, today’s research reveals that long queues and frustratingly out-dated payment systems remain a significant challenge that are ruining millions of shopping trips and therefore must be addressed. Despite the rise of online shopping, it is clear that retailers need to invest in the in-store shopping experience to maintain shoppers’ loyalty in today’s competitive market.”

Next’s ‘Kodak moment’ shows open goal for machine learning

Next is facing a ‘Kodak moment’ according to analysts at Berenberg, suffering from the effects of a ‘wait and see’ attitude towards the e-commerce shift which they were early to identify but slow to adopt.

Next’s strategy of pursuing short-term returns has left the retailer burdened with an outdated business model. Ongoing structural change in the industry means that retailers are set for significant gains from machine based learning which can help to predict future value from certain customer behaviours.
Pini Yakuel, founder and CEO of Optimove, comments: “Analyst predictions, technology innovations, competitor decisions – all of these form the landscape in which retailers have to create their strategy. Unfortunately, in the absence of a crystal ball, many decisions end up being based on current rather than future profitability. What we’ve seen is that application of new developments in machine learning can give predictions of future value based on individual customer behaviours.”
“For an industry which is still going through the challenge of some major structural changes, each of which comes with its own financial constraints, the power of machine learning to predict the returns on each type of customer interaction – whether that is a customer who only shops online, buys on their mobile whilst on the go, or who prefers to browse in store and buy online – is key to making savvy decisions.”

Related Posts