Customers need all sorts of information and help from the brands they deal with, and they’re used to expressing this as a question: “How do I..?” or “What’s the best xxx?”, with the right keywords in order. The search engines have trained us well over the years, to express a written – or increasingly now spoken – request for information succinctly.
Jeremy Payne
But as Jeremy Payne, Group VP Marketing & Alliances at Enghouse, reflects, it wasn’t always like that. Once upon a time if people wanted to know anything about a product they had to contact the company, by phone or even by letter – the idea of a third party service curating information about their stuff and providing it to the end user for ‘free’ would have been unthinkable. “Google really is the world’s most powerful knowledge management system”, he explained. “And they’ve done a lot right. When you go on there you feel confident that within the top 10 results you’ll find the information you’re looking for.”
This success in turn drives customer expectations when they query a businesses directly. But their ability to respond effectively really depends on how well that business has structured and organised their own knowledge base, for logical interrogation and retrieval of information
Sort the FAQ out
It starts with analysing the most repeated customer journeys in your organisation, the ones which drive the most profit for you and the ones which cause the most pain. Effective responses to these can probably be automated quite straightforwardly, because the response set is limited
Payne uses a metaphor of fixed exits from a roundabout to visualise this – you don’t need sophisticated machine learning to address it. He explains,
“When we engage with organisations we can see opportunities where a range of the self and automated service tools reviewed in the ContactBabel interactive self-service industry report would be valuable”
“But when you look at where they are on their journey and what their customers need, that’s a bit like first turning up to the Olympics and expecting to compete in the 100m sprint final”.
High-end automation
And while user expectations for personalisation of service do vary depending on whether they are calling their private bank or booking with a low-cost airline, even the most discerning customers are happy to receive an automated response, if it yields exactly what they need. Payne highlighted an automated email response service developed with their Eptica tool for luxury brand Montblanc International, using linguistic technology to generate replies:
“When they analysed the inbound queries they found that most people were asking about refills and suppliers for specific kinds of pen, so it’s easy to automate an immediate response which is clean and intuitive. Even for a premium brand, the speed and accuracy creates a premium user experience,” he explained.
Looking toward the future, a generation of digital native customers will expect even more from natural language processing to generate a single correct response, with the rise of voice assistants and smart search. We won’t care whether our device gets the answer from the business mentioned or from public content, we just want to know the most convenient place to get refills for our pen… Assuming we still use pens, from time to time.
But in a digital-by-design world, the real power will come from drawing the line between slick and accurate self-service, and the shrinking proportion of edge cases requiring the human touch. “Don’t try to automate everything”, Payne concludes,
“Automate where it makes sense… and use human nuance for the rest”
from UC Today https://ift.tt/2KI2HEl
0 Comments