Cisco UK & Ireland Blog

Programmable Retail: Writing Intentions to the Network (the Rainy Day Example)

April 25, 2017

The weather, what a celebrity wore, a winning recipe from a cookery show, an on-trend brand opening a flagship store nearby, long-term road or building works, a major sporting event, a royal wedding, a national holiday, a general election. All factors that can drive a spike or a slump in retail footfall and sales. The week of the UK ‘Brexit’ referendum saw a 4.6% drop in shoppers, while research has shown a seasonal temperature just 1°C higher or lower than average typically causes a 1% fluctuation in UK retail sales.

Some of these factors are (relatively) easy to predict and build into demand planning and forecasting, some distinctly less so and of course the impact varies greatly. Yet the Internet of Things (IoT) is opening up the possibilities for accurately anticipating shopper behaviour in real-time at a micro level, and directing in-store activity accordingly. With no crystal ball required.

Enter the world of programmable retail, or ‘algorithmic retail’ as coined by Gartner:

“Gartner describes algorithmic retailing as the application of big data through advanced analytics across an increasingly complex and detailed retail structure, to deliver an efficient and flexible, yet unified, customer experience. Algorithms connect big data to results.”, October 2016

Although algorithmic retail has widespread strategic applications, some of the greatest potential possibilities are at micro-level (i.e. in an individual store or warehouse). By fusing sensors (including smart phones and wearable devices) with machine-learning to ‘program’ an intelligent in-store experience, retailers can drive a seamless customer journey, and make the best possible use of resources. And with general and administrative overhead estimated as 15-18% of the cost of retail operations, Gartner highlights the value in optimisation.

At a micro level, programmable or algorithmic retail works in two main ways – either ‘writing’ (or pushing) primarily shopper intentions, or ‘reading’ (pulling) them, using a digital network foundation (for example Cisco DNA). The ‘writing’ element is most applicable to the customer-facing side, store operations, while the ‘reading side is applicable through the retail network and has significant supply chain benefits.

First, let’s take a look at how the ‘writing’ side of programmable retail works:

  • It’s raining heavily outside which is going to mean the polished floor of my luxury goods store gets very wet and very slippy, very quickly. A health and safety risk, for sure, and not the ambience my clientele is looking for. But a sensor in the lobby triggers notifications to store operations to ramp up frequency of cleans of the lobby area until the rain stops, when a notification is sent to return to normal frequency.
  • The rain means a large influx of people wanting to escape the elements. Sensors at the entrance and around the store calculate how many are displaying buying behaviour (lingering at displays) or heading straight for the café and send notifications that additional staff are needed to serve coffee or on the check out.
  • Many of the additional footfall are not regular shoppers – the guest Wi-Fi doesn’t recognise them from previous sign-ins, and they don’t have the app downloaded. The digital signs around the store show regular content highlighting the linked app and reward scheme and incentives for sign-up today.
  • And what about those that are regular customers? Wouldn’t it be nice to offer them a free coffee so they linger longer in store? It’s easy to push a notification when the app recognises they’re back in store. Or display an advert on the digital screens based on their online browsing history.
  • A customer wants to quickly purchase an umbrella but there’s a queue at the point of sale. Instead of putting it back, she can pay online via the store’s app (which conveniently reminds her via a direct message when she arrives in store). The SKU is then de-listed, meaning she can walk out without sounding the alarm at the security gates. This makes the Amazon Go model more practical for higher value goods.

Don’t miss my next blog where we will cover ‘reading’ intentions from the network – sign up using the box on the right. Or click here to learn more about how we are working with retail customers worldwide to securely digitally transform the customer experience, workforce experience and optimise retail business operations.

Leave a comment


  1. Si no dispone del tiempo preciso para efectuar estas tareas de limpieza
    siempre y en toda circunstancia puede contratar los servicios de
    estos especialistas que lo harán de forma eficaz y velocidad y con una calidad inmejorable.