Before lockdown, I was in London and I popped into the EE mobile phone store to add some data to one of my kids’ mobile phones. Speaking to the sales assistant, we got chatting, he checked my account, and told me that if I shifted to a different tariff, I could save some money

Pause for the catch. Was I obliged to purchase some other device to “unlock this saving”?

Perhaps there was some sort of elaborate insurance product that came with this saving?

Nope. Nothing. No catch. The huge swathes of data EE hold on me revealed my spending habits, and in a matter of seconds, gave the salesperson the nudge to help me reduce my monthly spend — I believe it was by £20/mth or something like that. Not a fortune, but enough to excite me and make me feel that not only was I getting better value for my money, but also that EE valued me as a customer, enough to challenge my spending habits and help me save money via an algorithm/data mining + sales assistant tag team.

Said tag team gave me total transparency about what money I could be saving, and built up the trust capital I had with them, in just a few minutes. Short term? They’ve lost £20/mth for a few months. Long term? They’ve made it virtually impossible for me to ever leave EE; I have no motivation to even consider another service provider, because the trust and transparency that led to significant cost savings has bought my loyalty for the coming years. They have locked me in and significantly raised the barriers for me to exit my EE relationship. And I’m not the only one; I suspect they are doing this for hundreds of thousands, if not millions of customers. Customer experience at scale, in such a precise and targeted manner, can only be done thanks to technology and machine learning.

Even if I were to come across an aggressive and competitive new contract offer from another provider, at the forefront of my mind, would always be the time they saved me twenty quid a month for doing nothing and how I was wowed by their customer service. They’ve created stickiness.

After the initial buzz of my new found wealth had died down, I started thinking more about the relationship between the data and the sales assistant. As a long-time customer, with a number of accounts linked to my name, EE holds a considerable amount of data on me and my spending habits. Their data revealed that I was spending too much for what I was getting; but seeing as I hadn’t questioned this, the easy and most profitable course of action would surely have been for them to take my money and carry on regardless. Like any consumer-focused company operating in a highly cutthroat, and in some respects, over saturated, sector, they instill sales KPIs for their customer-facing store assistants. That’s just how business is done, right? Profit goals, hitting monthly target, driving profit, sell, sell, sell.

Well, it seems that the people at EE have redefined how they synchronously extract from and deliver value to, their customers. The complexity of calculating long-term-customer value over immediate profit was quite something; only to be done via machine learning; a salesperson / human decision maker on their own wouldn’t have been able to come to such a conclusion and in such a short space of time. Saying that, I may not have been as receptive to the offer had it not been presented by an engaging human salesperson. Had the saving simply been an automated email, I maybe wouldn’t have believed it so quickly. But the human aspect to the transaction — that was an extension of brand EE and what they represent as a brand.

Instinctively, people are target driven and not customer-centric, especially in the financial services market. Weighing up and calculating targets versus customers’ needs is symbolic to an organisation’s leadership approach. Targets can be driven and delivered by data and chatbots. Customer service? Not so much. Yet. Bias and feelings in humans is normal. If I were to walk into an EE store, would a targets-driven sales person have given me a contract reduction without tangible data gathered from literally thousands of data points, come to the same conclusion? No. To be very clear. Alone, the salesperson would not have been able to make such a quick decision regarding my spending habits + long term viability as a customer of the company. Alone, without the salesperson, the data would not have engaged me or been able to deliver the news in such an engaging way that would implore me to stick with the company.

This entire scenario got me thinking more about how we’re building customer experience and automated, machine learning into our decision making at Sparkle. We’ve been fixated on building a chatbot, Indy, who we hope will be the ultimate hybrid of human personality + ability to be able to make unbiased decisions, at speed, thanks to access to thousands of data points. When we’re building and improving upon Indy, we even think about what her voice would actually sound like, were she to be in an equivalent position as my friend in the EE store. When developing her skills, we think about the senses; sight and hearing in particular — how does she make you feel after you’ve engaged with her? What senses does she evoke in you? We want to connect our ethos of trust and transparency, powered by data, with the human senses and reactions that we all still crave.

We’re used to human sales. We engage with people. We expect and assume people will make the right choice, based on their experience. Yet it’s becoming more and more obvious that the right choice is now powered by data, and with the removal of human bias. We’re championing the rise of the chatbots — that’s the only way customer experience will scale. We’re building a public chatbot that will be as pleasant and effective as a human, and if we get it right, I’d rather have a chatbot with an 80% human type experience that gets it right every time, than a 100% human that gets it right 10% of the time. We want Indy — her voice, her decision making, her aptitude, to be an extension of the Sparkle brand; to excite all the senses and to help our Sparkle customers feel the same way as I did when I left the EE shop