Today we use big data in every stage of the sales cycle. From the top of the funnel right through to purchase and customer service, we record trillions of digital interactions and mine them for ways to perform better. All this data is enormously useful because it makes the cycle visible and easier to understand. It has revolutionised the way we target customers as we have an almost end-to-end view of the interactions between us.
I say almost, because there is still one channel that we don’t have full visibility over: voice. Voice, even in our increasingly automated world, is still one of the most essential communication channels between brands and their customers – particularly if an issue is complex or urgent. While the content and tone of conversations between customers and sales teams or service agents remains invisible, we’re missing out an important piece of the puzzle. It’s a piece that, fully captured, can be applied to help our sales and service teams go from good to great.
Of course, we probably know who our rock stars are. We probably also know who our weak links are. There are also staff in the middle ground who are not bad, but also not great. With a little bit of guidance, they may make the transition to high-performing individuals. This is transformational for a company.
There are also other factors that impact performance that may not be down to the skills or attitude of the individuals on the front line of the sales or service desk. The question might be difficult, the answer complex, or things are happening around the context of the call that the sales or service agents just cannot keep on top of. This could be product changes, price changes, promotions or other external factors such as market moves or competitive pressures. This adds up to a tough job that’s fast, dynamic and real-time – it’s no longer a low-level, scripted role, but needs professionals who get the best out of every call. Those who do the job need help and those who manage them need insight to detect the gamers of the system, remove the weak, but also to scale best practice.
Solving the sales and service front-line challenges starts with automating and capturing the context of the call at a much deeper level, automatically. This means not only getting the electronic transcript (i.e. basic speech-to-text), but also getting the true meaning and understanding of the call using AI and NLP techniques. In this way, we truly connect the dots on how high performers get their results. We see their questioning techniques, identify them and link this to actions and insights. This can be shared with the wider team to help them to learn and to emulate.
Staff with a tendency to “game” the system can also be identified. Many of us have heard the “bullish” forecaster: “Yeah these three deals are all at 90 per cent for the quarter and I’m committing to the forecast.” Fantastic. However, connecting the forecasted opportunity to the context of the calls shows there is no mention of pricing and no mention of contract or GDPR, security, or timing. Therefore, the sales forecast can be checked against whether late-stage sales language is being used in the prospect interactions.
It’s not all bad process or bad staff. Assistance during calls is increasingly valuable. Sales and service professionals have a lot to manage during a customer call. Speech technologies are increasingly being integrated with external data feeds and internal knowledge bases so that some of the pressure can be taken away from the agent, sales or service professional. We are now working with some of the largest organisations in the world where speech analytics and conversational computing technology is helping to augment the capabilities of the human. We digitally listen to live calls, analyse them for understanding in semi real-time and integrate with internal systems or external data sources to present advisory actions to the agent. This dramatically improves first call resolution, sales opportunities and customer experience.
This is increasingly valuable in the multichannel world we live and sell in. Many omnichannel strategies are designed to solve one problem, but regularly create many others. The main issue is, does the left hand know what the right hand is doing? If not, this is immensely frustrating not only for the customer but also for the call handler, whether that be sales or service.
Being pre-briefed before taking a call is incredibly helpful. In many cases the agent does not need all the intimate details of every orders or all calls, emails and chats this week, on their screen three seconds before the call. That is almost impossible to digest, but what is useful is a summary that covers the volumes of those calls, emails, chats – is it the 20th interaction this week? That should be handled differently from the first – and gives an overview of the sentiment – is it positive or negative? It should also flag the topics and products detected and any complaints or negative language used.
That basic information gives the agent a pretty good understanding of the call context before they pick up the call. In fact, the data collected may also be used to impact routing or channels made available to the customer.
One thing is for sure, our staff are still a very powerful and capable resource. Now we are augmenting their capabilities with truly incredible technologies, but also giving our leaders the X-ray glasses so they can see what is going on for real and use that data to move the performance of their teams up the scale, with the goal of making every call an opportunity for the business.