Two Scottish businesspeople wanted to predict how much money to set aside for a business venture. Wisely, they turned to Edinburgh University for assistance. Based on the pair’s data and insights, they created a model that was just £1 out following launch.
That level of accuracy might surprise you. But what’s more surprising is the businesspeople in question were the founders of Scottish Widows – now worth billions – and that they reached out to mathematicians back in 1744.
People often talk about the ability to model the future in hushed tones, like we’re only on the precipice of such a discovery. But as Data Strategist Rob Jackson explained at a recent Rackspace and Google Cloud Platform roundtable dinner, this has been going on for centuries. Rob suggested, “The main change today is that technology is no longer a barrier. This presents a massive opportunity for retail, because there’s an immediacy to predictions due to the large number of data points and the sheer power of modern platforms.”
Yet despite all the promise and scope this presents for innovation and creativity, Rob noted that “many businesses fail to realise the potential of data. As few as 30 per cent feel they are data driven. Only rarely do organisations fully utilise data as a strategic asset, doing something tangible to improve the performance of their business.”
At the event, Rob demonstrated how a company called Programmai has been revolutionising data use for a range of retailers. The system – built on Google Cloud Platform (GCP) because of its scalability, cost efficiency and reliability – showcases what’s possible with smart technology, imagination, and the right partners. It works by introducing cutting-edge, deep-learning models to customer data. This can transform how a business operates in a manner that seems like science-fiction, but that’s also a logical evolution from the 18th century mathematicians mentioned earlier.
Rob’s first example of Programmai in use examined it being applied to a large retailer. They’d found customers would purchase once but never return, but those who bought a second time usually became long-term customers. Machine learning was used to predict who would only buy once, step into the customer journey, and target such people with promotions and online advertising to follow up on their purchase. They then invariably went on to buy again.
The experiment was extended to analyse behavioural patterns of anonymous visitors, targeting promotions to those with the potential to become lifetime customers. But also, anyone with a very high likelihood of spending wasn’t targeted at all because the model knew they were likely to come back anyway. This mix of cost savings and individual targeting resulted in a massive uplift in incremental sales that simply wouldn’t have happened without the AI layer.
Programmai also boosted the fortunes of a DIY retail client that often converts purchases over the phone. Predictions were made on the flooring samples people ordered, and a call centre experiment set up. Half ran calls in the normal way, the other half used a ‘hot list’ powered by Programmai to predict high order values. The average order value of Programmai calls was three times that of a standard call.
These examples are inspirational, but also showcase key insights into the future of retail. As Rob noted: “Disruption through machine learning affects all points of the customer journey and can be wide-ranging. You might start with a distinct business problem; but once you succeed, this opens other ways to be creative and communicate with the customer, putting them at the heart of everything you do.”
A major challenge facing modern retail is making the process of buying online easier than being in a store, and so such communication is vital. The likes of Programmai can be a virtuous circle – oil in the gears of the retail experience. AI powers an enhanced consumer experience, making it more informed for the user. It also notifies representatives in call centres and stores.
But modern thinking requires modern infrastructure. Several retailers at the event said they felt hampered by organisational obstructions like legacy systems and old-school IT companies. There was also concern about recruiting – and keeping – data engineers without moving premises. Some were also excited about the prospects of embracing the cloud and fully utilising data, but a little unsure of how to approach such a radical shift in their business.
As Rob explains, GCP can “help navigate around many of the obstacles you face, while enabling you to spin up prototypes and bootstrap innovation. AI will then work with the data you already have but aren’t making use of.”
As the first premier managed services partner for Google Cloud Platform, we can help you drive transformation, and experience the agility and flexibility the cloud offers without it becoming your core business. Our expertise and skills can complement what your organisation already has, without the need to make big changes.
Interested in moving to Google Cloud Platform? Take advantage of our limited-time free 30-minute consultation and $5000 credit towards your Rackspace Managed Google Cloud Platform infrastructure.