Episode 1: Introduction | Tethr Learning Series - The four "D's" of better B2C sales performance

Matt Dixon, Ted McKenna, Tom Shepherd

March 22, 2021

Talk to a product expert

This material originally appeared as part of our Learning Series podcast on B2C sales. Listen to the original here.

Welcome to the Tethr Learning Series, The four "D's" of better B2C sales performance. In this learning series, we're going to be going into a lot of detail around the research and some of the findings that went into our recent HBR article by the same title, co-authored by myself, Matt Dixon, Ted McKenna, and our colleague, Tom Shepherd.

In this first episode, we're going to recap the findings from the article, look at the what, why, and how of our research, and then ultimately explore the implications of those findings for sales organizations over the next few episodes. 

Back to where it all began...

First, let’s look at why we even decided to do this research to begin with. In the HBR article, we shared the findings from a large study that we ran. In this study, we built a predictive sales model, where we looked at B2C sales performance, specifically in inbound sales conversations. In layman’s terms, we’re talking about when customers call companies to buy things. So that could be B2C sales, it could be B2B, but it could also be as simple as calling a mobile carrier to get a quote on a mobile plan.

In the course of our study, we looked at more than a dozen companies, and we built a predictive sales model that contains more than 8300 independent variables. Then, we applied that model to a broader sample of more than two and a half million inbound sales calls. 

And what we found is that the best performers in B2C sales exhibit four key behaviors.

  1. Disqualify aggressively
  2. Drive customer decisions.
  3. Dig into customer objections
  4. De-risk the purchase decision

We found a huge return when salespeople exhibited these four behaviors. Conversion rates are several orders of magnitude higher than when they don't. In the following episodes, we're going to get into a lot more detail on each of these behaviors individually. 

Why do this research at all?

But before we dive in too deep, let’s talk a little bit about the occasion for this research. Thanks to COVID-19, the inbound sales call center has become a much more important channel, which is part of the story. But there’s more to it than that. This trend has actually been in progress for several years now, and changed the way companies think about this channel; long before COVID-19. Contact centers are being hit with a lot of pressure in terms of conversions, upsell, cross-sell, and more. Expectations for those agents may shift more towards sales or spotting opportunities beyond just service. We see a lot of these instances with our customers, and we help them pivot and deal with those challenges.

At Tethr, as a conversation analytics platform, we’re good at helping companies surface insights from unstructured data. It’s what we do. And we’ve seen a lot of fascinating insights come out of the inbound B2C sales centers recently, which is essentially why we decided to put together this project. 

How it works: Our research process

At the highest level, what we do at Tethr is take unstructured conversations and process them into something we can analyze. In this case, we're talking about sales conversations. So, a customer picks up the phone, they call their insurance company, their mobile carrier, or their bank, right, and they're talking to a sales agent—those conversations are recorded as audio. That’s our unstructured data, and we’ll put it through the following process to turn it into something we can use.

Start your free 30 day trial

Step 1: Generate unstructured text

The first step is to turn our calls into text. For this, we use a process called transcription, also called automated speech recognition. Now, some of the interactions in our study, were already texts, because those were chat-based interactions. So some of them were already text, but most of them that we looked at, were actually audio and we had to convert them. Now, at that point, we still just have a blob of text. It’s still a big mess. 

Step 2: Machine learning

In order to bring structure to this unstructured text, we use something called machine learning categories. Think of a category as a training set; it is us giving the machine instructions to go find when a certain thing happens in a conversation. For example, let's say, we're looking at auto insurance sales phone calls, and we're looking to find what percentage of customers are calling in for the military discount program. In this case, we can create a category which contains all the various ways in which customers mentioned the military discount program. The machine can then spot all the times that “military discount program” is mentioned by the customer or by the agent. 

This is a very simplistic example, and we can certainly use more complex techniques, including different demonstrations and avoiding false positives, but this is a good example of how categories function. Each category is carefully built, trained, and tested as we work to train the machine to know what to look for in each one. 

Step 3: Derive insights from that data

At Tethr, we have more than 1000 out of the box categories that we use, which is how we start to bring structure to this unstructured data. We have a wealth of transcribed call, chat, and case data, now all tagged with insights. We know when those insights occurred, when those things happen, how many times they happened, in what sequence or maybe in what reaction to something a rep may have said or may not have said, etc. Now, we look at how all that data pertains to the outcome of the sale.

This process makes it possible to track the behaviors that statistically drove the outcome that we care about: a closed sale.
Download BCLC Case Study

We looked at these different insights, these different events, in conversations, and how those things drove or depressed sales conversion. We also had a known outcome for every single one of those sales conversations in our study.

This machine and this process give us really, really powerful ways to point to very specific agent behaviors and say, we can definitely know that these behaviors make a big difference when it comes to conversion.

The most surprising takeaway from this study

We learned a lot during this study, but one of the most surprising takeaways was this: Think about today's buyer. There are so many easy ways to buy online. There are so many easy ways to go and purchase. And yet, these callers are people who have chosen to not do that. If your buyer can purchase a pair of jeans online, but they still walk into the store to go try them on, there's a lot of lessons about understanding who that person is. 

What type of buyer foregoes the “easy way” of purchasing through your website, and instead chooses to talk to a human to conduct that transaction?

None of those B2C sales calls occurred with products or services that customers could not have bought online. But nevertheless, the customer called in. Why did they call in, rather than purchase online? That’s the question of the hour. Maybe they’re stuck, or they don't know if it's the right decision, they're stuck between options or looking at Plan A versus Plan B, but there's something and they want to have a conversation in order to make that decision. That’s a fascinating look at the type of customer who calls into a sales center in the first place, in this day and age.

The payoff

None of this stuff really matters if it doesn't drive sales conversion. We’ll talk more about this in later episodes, but the big takeaway is this: If you get the majority of these high-value behaviors on a call, you're getting close to 60% win rate, potentially all the way up to 75%. So real huge lift potential by doing all these things in conjunction with one another. Just going from not that great to average or even a little above average gives you a significant lift in terms of win potential.

If you get the majority of these high-value behaviors on a call, you're getting close to 60% win rate, potentially all the way up to 75%.

In conclusion...

The truth is, most of our sellers are already practicing one or more of these behaviors. Even if it’s accidental, we're not really starting at zero in most cases. While it's probably unrealistic to expect that we would get all four of these demonstrated and used all the time, the lift is still massive. Some companies that we studied are fielding literally millions of inbound sales calls, and every call is seven to 10 minutes. Some of these calls are longer, some are shorter, but either way that's real money, right? These are real people that you are paying to convert those inbound contacts into closed deals. If you can improve your ability to swing those conversations by getting agents to do a little bit more of the good stuff, and a little bit less of the bad stuff, that can result in an enormous improvement.

Thanks for joining us for this high-level walkthrough of the what, why, and how of our latest research into inbound sales. If you enjoyed this episode, make sure to sign up for our Learning Series to stay up-to-date with upcoming installments. Join us next week as we take a deeper look at the first of the four behaviors that drive B2C sales!

This material originally appeared as part of our Learning Series podcast. Listen to the original here.

Download The End of Empathy Ebook
Jump to:

Most popular articles