動画
With more companies asking for feedback and response rates under pressure, advanced analytics can help businesses infer customer feedback from a growing volume of digital interactions. Tom Springer, a partner with Bain's Advanced Analytics practice, shares how companies can use descriptive, predictive and prescriptive analytics to understand implicit and explicit customer feedback, and act on it in real time.
Read the Bain Brief: The Future of Feedback—Sometimes You Don't Have to Ask
Read the transcript below.
TOM SPRINGER: How do you achieve customer intimacy at scale? It can be really difficult with hundreds of thousands or even millions of customers. How do you know what each one is thinking and feeling, and how do you transmit that information to the hundreds or thousands of frontline employees that you're expecting to deliver personal and attentive service?
You certainly can and should ask for explicit feedback, right, through customer surveys, but this is necessary but not sufficient in today's world. More and more companies are asking for more and more feedback.
Response rates are under pressure. It's not a given that every customer is going to respond to these surveys. And so how do you find out what's going on with those that don't respond?
Advanced analytics can help by inferring customer feedback from the vast volume and growing volume of digital interactions that are happening in most businesses today.
Amazon and Netflix are great examples and well-known examples of companies that do this well today—tailoring their experiences based on inferences they draw out of customer behavior. But Amazon and Netflix don't have a monopoly on this capability. Every company can develop the analytical capabilities to infer customer feedback.
We break these capabilities into three categories: descriptive, predictive and prescriptive. Descriptive analytics literally describe what happened before, during and after a customer interaction, and companies can use this information to develop insights and potentially new ideas about how to treat situations that often create customer delight or detraction.
So for example, if a flight is late, typically customers of an airline don't like that. And so if you observe what happens, you might be able to actually change just systematically the treatment of late flight arrival.
Predictive analytics take this a step farther by inferring what's likely to happen based on history. For example, one telecom company found that when a customer defected, that others in that customer's calling circle—friends and family—were at elevated risk of defection also. And so their predictive model helped them flag those situations as they were happening and trigger actions based on those predictions.
Prescriptive analytics take it even one more step, which is to say, automatically detecting risks or opportunities and then triggering an action based on that detected situation. And so a good example of that would be, come back to Amazon or Netflix, if I'm browsing a different movie category on Netflix than I was before, maybe new recommendations need to be served up to entice me to watch a particular movie and continue being engaged with Netflix.
Working together, these capabilities enable companies to understand implicit customer feedback, along with explicit feedback, and act on it in real time.
We find that companies approaching this task find it daunting and are best served by an evolutionary crawl, walk, run approach. Starting small, building up to more complicated and extensive use cases over time.
So how do you get started? Well, start by asking some basic questions. Where are we on the evolutionary scale? What experiments might help move us forward? And how do we increase the speed of iteration and learning?
For most companies, this will be a multiyear journey, but one well worth taking on the way towards having a customer franchise that will be the envy of the competition.
Read the Bain Brief: The Future of Feedback—Sometimes You Don't Have to Ask