Taking self-service to the next level

Having attended Nuance’s annual Customer Experience Summit in London earlier this week, it’s clear that self-service automation is making real inroads.

Nuance reported that it now processes some 14 billion customer interactions a year, and that is translating into significant business benefits in terms of an improved customer experience and cost savings for service providers.

Another key theme of this year’s event was customer enthusiasm for well-designed self-service applications. For example, when Domino’s Pizza trialled its Dom virtual assistant, it handled 200,000 orders while still in beta.

Three specific takeaways from this year’s event:

• Voice biometrics for authentication will quickly become the de facto way of identifying and verifying customers
• Blending voice biometrics with big data analytics to deliver new levels of insight into fraud
• Taking Virtual Assistant deployments to the next level with full conversational interactions

With HSBC recently joining Barclays in offering voice biometrics authentication, there’s now a major proportion of UK banking customers that will be using the technology. I believe it will quickly become the standard for access, and that customers will begin to ask why their other service providers aren’t protecting them with voice biometrics. Barclays talked about how those customers that don’t go through biometrics still have to go through ‘the quiz’ – a tedious verification process. It’s easy to forget that it wasn’t very long ago that we all had to go through this for every service interaction, and it just shows how much momentum there now is behind the technology.

Barclays also discussed its Barclaycard fraud mining project, where it’s entire database of call recordings was matched against voice biometric prints. Doing this they were able to identify fraudsters with multiple accounts, indeed they isolated one criminal with almost 50 different accounts! Cleary this kind of exercise requires a significant commitment in terms of processing power, but the effort is worth it if the result is that fraudsters no longer bother with Barclaycard as they know they’re likely to be picked in near real time. Unfortunately, what’s good news for Barclaycard isn’t always good news for other financial services providers, as those fraudsters thwarted by one firm will simply move to the next target. I suspect we’ll see a lot more biometrics-powered fraud initiatives over the coming months.

The Nuance event also highlighted how Jetstar – the low cost airline from the Asia Pacific region – is putting intelligent virtual assistants to work. The airline’s Ask Jess virtual assistant has moved beyond first generation VA deployments – that were often little more than front-ends to a knowledge base – and delivers an impressive conversational performance. It’s like having a webchat with a live agent, and this is clearly the way forward for virtual assistant deployments.

Jetstar also showed just how flexible the technology can be. In Japan, for example, they’ve moved away from their Ask Jess persona to use a friendly bear as their customer persona – showing the technology’s ability to be localised for different markets. This also highlights the need for effective design and dialogue skills when it comes to developing and deploying self-service solutions.

Nuance also emphasised the key role that machine learning will play in its next generation offerings. A preview of HAVA, Nuance’s Human Assisted Virtual Assistant showed how with ongoing machine learning automated virtual assistants will soon be able to front the entire customer conversation, but with the intelligence to reach out for live support when it becomes necessary.