Whether it’s chatbots, intelligent assistants, natural language processing, machine learning or big data, it’s hardly surprising that much of the Conversational Commerce debate centres on technology. However, even the best Virtual Assistant solutions require ongoing optimisation if they’re to succeed for organisations and their customers.
With less than two weeks to go until Opus Research’s Conversational Commerce Conference in London, there’s lots of interest in the new technology components that are driving the next generation of conversational commerce. But, as Dan Miller – the lead analyst and founder of Opus Research – outlined at Sabio’s recent ‘The Art of CX’ conference: “the upcoming era of true intelligent assistants will require real UX expertise and processes if they’re to really deliver for customers”.
Here at Sabio we understand this imperative, particularly in terms of how an intelligent assistant deployment will be used to unlock value. This might involve integration to deflect and triage live chat, being used to support website interactions or grow channel rich, or just to support better managed transitions to human assisted service in the contact centre. It’s only when these goals are clearly defined that you can start to develop and optimise your intelligent assistant strategies.
Need for ongoing optimisation
Working across a range of virtual assistant deployments, we’ve found that organisations achieve the best results when they commit to investing in the ongoing optimisation of their intelligent assistants. This may take the form of a range of post-deployment services, from Conversation Reviews and ongoing reporting, through content and grammar management as well as training and consulting around conversational content writing, improved UI design and experience optimisation.
That’s why CX teams need to apply the same levels of process governance and analysis to their Virtual Assistant deployments as they do to existing speech or text analytics activities. Key to this is an ongoing assessment of all interactions to learn which ones could now be handled by an optimised assistant, as well as identifying areas where assistants need to improve. Speed is of the essence here, with ongoing reviews aimed at improving existing content, identifying areas that are missing, and fine-tuning of language to improve the customer experience.