On Wednesday 23rd May I hosted an event on Artificial Intelligence and Retail. We heard from 4 different presenters who each have extensive experience in the AI space.
The presenters were:
Anna Bravington - Director of Retail at Tide - Connected Commerce
Daniel Moisset - Founder of Machinalis
Maria Flores Portillo - General Manager at Persado
Karina van den Oever - Principal at Elixirr
With the focus being firmly on the practical business side considerations, and less about the technology itself, these were the key takeaways:
Artificial Intelligence Is Already Here
There is no point debating whether or not this technology is going to have a major impact on the industry, it’s already here, and it’s already helping to separate the more agile retailers who are investing in technology from the slower competitors who failing to grasp its potential.
In fact, this concept, of breaking down any type of human thought process into its component parts so that it can be replicated by a machine, has been around since the 1950s.
While Pepper the Customer Service robot is very cute, the reality is that AI is rarely about humanoid robots. It is often most effective when it is used to automate very specific, simple tasks such as creating segmentation within lists of customer emails, and trialing different marketing campaigns. And no, there are not any signs just yet of robots taking over the world. What we have now is "narrow" or "weak" AI where the computer can learn to do a simple task, or number of simple tasks combined.
AI technology is touching every part of the retail industry. It has back-end applications such as supply chain transparency, warehousing automation, stock forecasting and allocation optimisation.
For marketing it can be used to segment campaigns, personalise marketing communications and improve customer profiling.
In stores, facial recognition can now be used for payment in China, and AI can monitor queues and improve customer flow round stores.
From a customer service point of view, conversational chatbots can minimise the number of customer queries that have to be dealt with by customer service agents, and can even entertain customers or make recommendations.
In the eCommerce domain, AI is creating better recommendations for customers through learning about their preferences, or even what kind of clothes will suit them by using image analysis. It can allow customers to find goods based on an image they upload, using image search to improve the results that they show the customer.
The applications of AI technology is endless - it is almost as if AI has become something more universal, like electricity. AI can be used to power an almost infinite variety of applications that allow better, faster analysis of data and smarter decisions.
In addition, with many of these applications being cloud based and with the cost of running and implementing them dropping, smaller organisations will be also be able to benefit from AI.
Many of the applications listed above are analytical and logical, but “unsupervised AI” has the potential to be more creative, and look for solutions that are not so linear. AI can also be trained to recognise and react to human emotions. This is useful for example with chatbots, also with tailoring marketing messages based on the types of emotion that customers respond to.
Clearly, AI will allow retailers to make significant improvements in their speed of analysis, the quality of their decisions and their ability to offer personalised service to an increasingly demanding customer.
However, implementing AI depends on good data, and legacy retailers with vast amounts of data in silos across the organisation are going to have to commit to a significant operation to clean it up and get ready for AI implementation.
Many retailers are also not set up with the combination of technical and business knowledge that is needed to manage an AI implementation project, and some of them simply won’t know where to start.
Part of the challenge of AI implementation will be retailers taking the time to examine all of the AI applications that are available and identifying which ones will give the biggest return on their investment of time, resources and money.
Who is using AI well? Ocado announced last week that they had signed a deal to build between 3 and 20 automated warehouses for the US grocery firm Kroger, a move that saw their share price soaring. Ocado have long worked on their technology, which incorporates AI, to build their competitive advantage, and have made it clear that they are very much a technology and logistics firm as much as a retailer.
Another retailer who considers themselves a technology company at heart is ASOS, who also use AI to power their search and recommendation engines online.
Aside from these few high profile cases, many retailers have been slow to embrace AI. In a large part due to their issues with siloed, inconsistent data, AI has been lower down on many retailers priority lists. However, given the competitive edge that it lends to nimbler, more tech-savvy competitors, this approach may well be woefully short-sighted, if not fatal for some.
Retailers who want to remain competitive would do well to consider where they can use AI to streamline their stock, remove stumbling blocks from their processes, reduce manual labour in a time headcount reductions, and delight the customer.
For our next event, we will be looking at how retailers can use their brand, and their brand story to connect to customers. With retail continuing to be a turbulent place to be, how can a clear, compelling brand story help retailers of all size connect to their customer and drive sales?
Tickets are available here:
I hope you will join us for what promises to be another fascinating discussion.
We just sent you an email. Please click the link in the email to confirm your subscription!