Discussing the Role of AI in Customer Service with Simon Yencken

Yes, so it’s interesting. We started really focused on retail e‑commerce, and then we found other verticals were interested. We actually got our first airline account, Alexia, about six years ago. At the time, we built the company thinking the opportunity was retail e‑commerce, and here was an airline expressing strong interest in working with us. You might wonder what the connection is. If you’re building a retail site to sell fashion, how could that be similar to an airline? But data created by visitors is essentially the same. You’re tracking and understanding visitors: what’s their intention, and how can you provide a better service?

With an airline like Calitaria, the challenges are different but similar to selling women’s clothing. They have an inventory of flights they want to fill. When people look for a flight between Milan and Rome, or search from Rome to England six months out, that tells you about their intentions. Understanding those behavior differences lets you push people to add a rental car with a flight or bundle a hotel, selling more seats and filling flights with excess capacity. We were surprised but thrilled to see that behavioral data is important in every industry.

We work with both high‑end and fast‑fashion brands. A high‑end example is Dolce & Gabbana, a classic Italian fashion brand present around the world. They’re not interested in anything inconsistent with their brand image. Fast‑fashion accounts focus on selling inventory and conversions. In the luxury market, you don’t want to discount your merchandise. With a client like Dolce & Gabbana, they’re more interested in personalization: collecting emails and contact details more effectively, getting more subscribers, and if someone is looking at an outfit, suggesting sneakers, a belt, or a handbag to go with it. Those are the kinds of services Fanplayr can enable by understanding behavior and providing a more personalized experience.

Another completely different vertical is the telco space, for example Vodafone. We have many telco clients interested in getting more subscribers to a mobile phone plan, selling more mobile handsets, or selling home internet. Increasingly they might resell a service like streaming. Behavioral data helps them understand visitor interests, stop people from leaving by engaging them longer, and entice them to sign up for a mobile plan or a streaming service.

We found that in any vertical, businesses working with Fanplayr leverage behavioral data to create a better, more engaging experience for their visitors and ultimately make more revenue.

I think one thing I’d like to do is ask people to tell me about AI. They say they’ve never heard of it and ask if it’s real or just robots and scary machines. Then I ask, do you use Amazon? They say yes. Have you ever seen Amazon saying “You might like” or other suggestions for you? They’ve seen that—that’s AI. During the lockdown, have you ever sat in front of the TV and watched Netflix? Have you noticed they recommend programs you like, but if it’s your son or daughter, they recommend different programs? I’ve noticed that—guess what? That’s AI. When people start to understand that some of the iconic brands in this decade make heavy use of AI, they really start to get motivated to think about AI in their own business.

One piece of advice would be to dig in and make a start and not feel you need to hire data scientists or rent coffins, because there are some pretty good solutions you can try to see how they work. Product recommendations is a great way to begin.

We partnered for our AI recommendations with Google and use Google AI services. We found that to be incredibly powerful. We benchmarked it against competitive services. People already doing recommendations moved to Fanplayr AI and said, “You’ve improved our results by 200 percent,” so it’s outstanding.

When we started, we first launched product recommendations in Japan because we had many fashion accounts there keen on adding recommendations. We got some press and announced it to the market. Then Google contacted our team in Japan and said, “What are you doing? It’s Fanplayr powered by Google AI,” and at that stage they hadn’t launched it yet. It’s a great partnership and very easy to use off the shelf. I’d say any business should add product recommendations, because that’s AI that can really improve your site experience.

There’s some interesting trends. The one that comes immediately to mind in the news is AI and cyber security because we keep hearing about smart hackers. They say they’re from Russia—I don’t know if that’s true—but they’re breaking into companies and government agencies. If you think about the sophistication of hacking, it’s incredible. A lot of science goes into it, and AI can help anticipate threats and vulnerabilities. It can also detect unusual actions and flag them as potentially malicious. For me, that’s a big trend.

Another trend is conversational AI in customer service. Chatbots and virtual assistants are getting better at providing good service before you need to speak to an individual. Using conversational AI to deliver better responses to service requests is already emerging and will be a major development.

A third area is the fusion between AI and IoT. Think about all the smart devices we have now—security cameras that let people in, home thermostats, and many internet‑enabled gadgets. You might say, “What is Simon talking about?” But imagine adding AI to make them smarter and more intelligent, anticipating things like home heating and lighting. Right now, with IoT devices and platforms like Apple HomeKit, you can set timers so things turn on when you get home. But imagine devices that recognize you’re on holiday and turn the heating off, or notice that you want lights on at a certain time. Or adjust heating not just by ambient temperature but by behavior patterns. That fusion of AI and IoT could be massive.

Finally, an important trend that will transcend all of AI is ethical AI principles. Some commentators, like Musk, warn that AI could be evil and machines could take over the world, which is scary. But there’s also the issue of fairness and diversity—ensuring AI treats people equally regardless of background, ethnicity, or income. I think we’ll see more pressure on AI designers to incorporate ethical principles into their technology.