DVELP-mobilearrow-flex-shortarrow-flexarrow arrowsbinocularschartcogsflex-icongeargithubglobeheadphoneslightnings linkedinlock newsletter-chartDvelp_Newsletterplantscalestarget triangle-icon twitter

Leveraging Machine Learning Into a Retail Marketing Programme

December 10, 2020

When we talk about AI today we are really referring to Machine Learning (ML). ML is a triumph of pattern recognition. It is amazing at processing vast amounts of data to find patterns that humans could never spot and make predictions based on those observations.

ML was once the preserve of the silicon valley giants but the democratisation of this technology through the cloud has made it available to a wider audience in the last 12-18 months. One example of this is the application of ML to speech recognition and natural language understanding which can take on the role of concierge and handle part or all of the conversation with a customer in your contact centre, whether it be on voice, messaging or any other channel.

At the Retail Bulletin Board’s ‘Future of Retail Marketing’ digital event Stuart Dorman from Sabio was joined by Geoff Bull from Swyft Sofas for a panel moderated by Darren Williams from Williams Harding Consulting. For those of you couldn’t make it, here’s a 3 minute read summary as well as the original recording.

Good implementations of AI into your customer service environment should be simple, easy-to-use and non-intrusive. They should also be used as one feature of a wider technology infrastructure designed to help the customer progress to solving their problem. Since the birth of the call centre industry we have strived to reduce friction from the processes customers have to endure to resolve their problems. Back then the Yellow Pages were widely used as a primary route to a business, resulting in the first quantum leap: using the telephone as your primary user interface to an organisation.

A few decades later a second quantum leap occured; we engaged with companies through graphical user interfaces by way of the internet. Though it opened up opportunities for self-service, this new interface was less effective at solving complex queries, and so in cases when customers had a complicated enquiry they would revert back to the phone. The conversational user interface provides the latest quantum leap to organisations looking to bring these technologies together and digitise conversations.

Much the same as the telephony user interface augmented physical stores and the graphical user interface augmented the telephone, Conversational AI should be seen as complementary to the user interfaces that came before it. In the form of a Virtual Agent it can either solve the customers problem entirely or - more commonly - can handle some of the enquiry and then pass the customer to a human agent for the complex parts. The key to success is to recognise and understand how we as humans communicate. We speak faster than we type; we read faster than we listen. We are able to absorb vast amounts of information in a split second through images at a level of fidelity far greater than with words. However, the precision of language cannot be matched. We must design experiences that take advantage of the strengths and weaknesses of each user interface and ensure that experiences flow seamlessly between UI’s as well as from digitised channels to assisted channels.

The most successful deployments of Conversational AI start small. They first develop a handful of ‘customer intents’, or types of reasons for calling, and grow organically over time. Often people feel overwhelmed with AI and don't know where to start. It doesn’t have to be this way.

With this newly democratised technology we can get up and running and train a model within a few weeks. All it takes is around 25k calls to get going. We are helping organisations to achieve some seriously big wins with this approach.

Make sure you’re subscribed to our newsletter to be notified about future events!


By using this website you agree to our cookie policy