Brian Kenny: "A B C, always be closing." Film buffs might recognize that quote from the 1992 film, Glengarry Glen Ross. It's an iconic and agonizing scene where Alec Baldwin berates and belittles his sales team in an effort to motivate them, which he ends by saying that half of them will be gone by the end of the week. Not surprisingly, this method of boosting sales team performance is likely to meet with mixed results. Fortunately, most sales organizations take a more sophisticated approach, but it's not always clear what's working. The CRM marketplace is valued at $120 billion, but despite the huge investment many firms have made, the impact on sales productivity has been difficult to quantify. And let's face it, most sales teams have a love, hate relationship with CRM systems. They just want to talk to customers.
Today on Cold Call, we welcome Professor Alison Wood Brooks and case protagonist, Amit Bendov, to discuss her case entitled, Gong: Resonating Conversational Insights. I'm your host, Brian Kenny, and you're listening to Cold Call on the HBR Presents network. Alison Wood Brooks studies the psychology of conversation, why we say things we shouldn't, and don't say things we should, and she created a course in the MBA curriculum called, “How to Talk Gooder in Business and Life.” I love that title. Amit Bendov is the CEO and co-founder of Gong. And he’s the protagonist in the case we're discussing today. Thank you both for joining me. This is going to be fun.
Alison Wood Brooks: Brian, let's go. Thanks for having us.
Amit Bendov: Yes, I'm excited to be here. Thanks for inviting me.
Brian Kenny: So I think people will enjoy hearing about this case. I know we have a lot of listeners who are in the sales role. One of the things that HBS has been sort of taken to task on over the years, is that we really don't do enough to talk about the sales function and to help people understand the sales function. So I think this case it does that, but it also does it in a very interesting way. Alison, I'm going to ask you to start by telling us, what's your cold call to start this case in the classroom?
Alison Wood Brooks: Oh Brian, I love it. Before we even get to the cold call, let me give you a little bit of context about this course, “How to Talk Gooder in Business and Life,” because in a way it's a bit of it sort of mirrors what Gong aims for in that we try and get students... It's not a typical case-based course. They do a lot of practicing doing conversation, listening back to themselves and then trying to do it better or gooder. So for this case, before we even get into the classroom, in addition to reading the case itself and learning about Gong and thinking about Gong, I also ask them to do an exercise. The first piece is, think back to a recent conversation that you had. And I ask them to try and remember it and not just remember it, but can you rewrite it like a screenplay? Do you remember who said what in what sequence and how they said it, how you felt, all of those things? The second piece is, now, go ahead and actually record a conversation and listen back to it. So by the time they get to class, they've done this recall exercise, what can we remember from our human minds, versus what does technology help us hear when we listen back to an actually recorded conversation? So then when we arrive in class, the cold call is: does recording our conversations, help or hurt?
Brian Kenny: Interesting. That's a great way to start the class, a little nerve-wracking too, I would think. So. I mentioned a little bit about your course, which I do love the title. I want to know why you felt that Gong was a good case to write and how it relates back to the kinds of questions that you look at as a scholar.
Alison Wood Brooks: You wouldn't just like the title of the course, you'd like the content too, Brian. I'm a behavioral scientist and all of my research focuses on trying to understand how people tend to converse with each other, so describing all of the good things that we do when we talk to people and all of the mistakes that we make and sort of deeply understanding what we do and why we do it? And then, also using experiments to try and uncover strategies to help people do it better. And so, this case was such an obvious match. And I think Amit feels the same way. That's what Gong's trying to do too. They're trying to figure out what do great salespeople say and do? And can we take their strategies and help other people learn to do those things as well? And so in a way, Gong's mission is very aligned with the mission of my research as a behavioral scientist and of the course. I love starting my course by talking about this case and the sort of alignment between our two missions.
Brian Kenny: Great. Amit, let me turn to you for a second. First, I just want to know where the name Gong comes from. Can you give us the background on that one?
Amit Bendov: I looked for something that is for a short, does not have anything to do with the subject matter directly, but it's more evocative. And I thought about Gong. Gong's kind of like the sound, it's the audio of winning. When sales team, not in a Glengarry Glen Ross environment, but in modern, when someone just brings in new customers, they usually hit the gong. So this is it, it has a verb potential. You could say, "I'm gonging this conversation." And equally important, it was affordable. It was $1,000 from an Australian guy. That was it.
Brian Kenny: That's perfect. I like that. I liked the audio, it's sort of reference to the gong, that makes a lot of sense. Can you tell us a little bit about your background I think first? And then also, what led you to this idea?
Amit Bendov: There's a long story, but Gong is the fourth company that I'm leading. When I was a teenager, being a CEO in tech definitely was not part of the plan. I wanted to be in a rock band. And I actually started my career in recording studios. I was selling audio equipment to a recording studios and concerts. I signed up for a computer science in Tel Aviv University. Just for the listeners, I was born and raised in Israel. And that's where I started. Third year in campus I saw this ad on a software company that in an artificial intelligence multimedia. I wasn't sure what multimedia was, but I thought this must have to do something with show business, which is that's perfect for me. That company was actually called, ClickSoftware, recently sold for $1.5 billion, but I was one of the, I developed the product, one of the founding teams and I did it in Boston, spent a lot of time around your campus in Harvard Square. And then, we moved back to Israel, joined another company in the ERP space. And then I got a CEO job, my first CEO job was at a business intelligence company called, SiSense, the company started in Tel Aviv was a small company. I was brought in to turn it around. And that's where I got kind of the idea for Gong. It wasn't called Gong. And I wasn't even thinking about starting a company, but I realized that how much information we're losing. That in fact, we don't know anything about our customer experience and our people experience.
Our customer support people and our salespeople, we're having thousands of conversations a week with customers, but what would make it back to us as a leadership would be very, very short summaries. 99 percent of the information gets lost. Just physically, there's about a 6,000 words per hour in a conversation. If you look at 150 words per minute, what ends up in the CRM. CRM is the systems that people actually use to take notes.
Brian Kenny: Customer resource management, just for those listeners who aren't familiar with the term.
Amit Bendov: Think of it like a Rolodex on steroids. And we'll do 30 words. So 99 percent never makes it there. And the 1 percent is obviously highly subjective. It's what I understood that happened. So there is one loss already, because it's very easy to miss hear and misinterpret it. I mean, Alison, you spoke about the value of recording. And I thought about a system that actually captures information directly without asking people to enter anything and using natural language, understanding to capture all that information. That's how it all started.
Brian Kenny: Yes. So let me ask you, because you've managed sales, the sales function, you've managed sales teams. What's that like? What are some of the challenges that you encounter as somebody who is managing and trying to motivate and trying to understand, what the sales team is doing?
Amit Bendov: It's a combination of joy and agony. And sometimes it could be on two different, just on the same day. We deal mostly with things that are kind of complex. The salespeople, some of them are very successful, some are not. And we started to understand why, it could be lots of things. So usually, they'll put a field in a CRM system, like loss reason, if they don't win a deal. It'll be like no budget, no decision, they're ghosting us. It's never about what I could have done better. Second, you mentioned the David Mamet play. It's just like, okay, "The leads are weak." The cards that I’m being dealt aren't good. And it's hard to say, I mean, usually people don't believe them. If you remember what the movie like, "The leads are weak, you're weak." But sometimes it's true. So I realized that we don't know anything really about what's going on and we can't rely on self-reporting for the reasons that we've mentioned to know. So, all we have is guesswork, beliefs, anecdotal evidence, hypothesis, books. There're I think, 11,000 books on Amazon, last time I counted on how to sell. Very few of them are based on scientific study and evidence and data. It's all opinions and about experience at this company or that company. We're at 2015, Google just had software beat the world champion in chess. And then, AI started to diagnose a cancer better than doctors. So can AI understand conversation better than salespeople? Probably yes.
Brian Kenny: Let me pause you there, because I want to go to Alison for a second, because I have a more fundamental question than that, which is if the salespeople aren't accurately reporting all the details and maybe fudging the reasons a little bit, why something didn't go the way they wanted it to, why on earth would they want anybody actually listening in on what they've done? That would to me, that that takes away all my wiggle room.
Alison Wood Brooks: Yes. So the first part of what you said, I hear what you're saying. I think it's less about the salespeople fudging it. So the problems with self-report are multitudinous. It's both what you said, which is like "They're fudging it," but it's actually mostly we trust people and think they're good. They just don't remember. Or they don't have good access into causality. They don't know what they've done, they can't remember what other people have done, and so there's just no knowledge at all of what's going well and not well? And that's where the data like Amit is saying is quite helpful.
To your question Brian of, if I am in this environment that's hard and complex like Amit was saying, and maybe I'm not doing great, why would I want someone capturing that? And that's why we start the conversation in class of this case at this point of how does recording conversations help and hurt? And it raises some sort of, deeper, more complicated questions. One of the questions that we debate in class is, do people really want someone listening in on their conversations? The first part of it is, well, who's listening? It's really valuable for you yourself to go back and listen. Part of the problem with conversation and why it's so hard is that we don't get to listen to the game tape and we don't get good feedback from others. So we're just not getting any feedback from anywhere to figure out what we're doing well and not well. And so, having it captured so that we ourselves can go back, so salespeople can go back and listen to themselves and look at data analysis of what they've done, that is so valuable. And most of my students and most people feel this way. They see this very quickly. It's like undeniable evidence of what you're doing well and not, well. It's also mortifying of course, to listen to yourself.
Brian Kenny: Kind of like listening to yourself on a podcast.
Alison Wood Brooks: Oh, yes. I even, I have an identical twin sister. So I have spent most of my life feeling like I'm listening to a weird, bizarro version of myself and I still find it mortifying to listen to my own voice on tape. But then the concerns about recording conversations become a little trickier, when you think about who else is going to listen? Who else is going to have access to this data? Is it my peers? And of course, there's lots of learning you can do, if you're going to share data with your peers. If you have a friend who's killing it as a sales person, don't you want to listen to what they're saying and doing and learn from them? But also again, demoralizing, and that's hard. Do you want your manager to see you at your best and your worst? I mean, in theory, yes. But in the experience of it, that could be hard.
And then, when you think about who else has access to this data and who should have access to this data, then things would get much trickier with data privacy and the balance between what can we learn versus what is ethical. What should we have access to and what should we not have access to? And those privacy concerns are relevant for Gong's business, they're also relevant for researchers and scientists like me, that I think the world will be grappling with more and more, as more and more data is available.
Brian Kenny: Let's just talk about CRM for a minute, because many people listening probably are in firms that have spent millions, tens of millions of dollars implementing Salesforce and other kinds of CRM systems. I teased in the intro that the returns on that are somewhat questionable I think, for a lot of these places. And I think, salespeople have shown kind of mixed emotions about CRM systems. And it's a lot of work. It's a whole change management process. I really don't want to enter all this information. Amit, I'll start this with you. What do we know if anything, about the kind of efficacy of CRM systems?
Amit Bendov: These are important systems. They're not wrong. And they play an important role, I mean, at the very basic, you need to log who your customers are, what they bought, what their contract is. So there's got to be somewhere to store all that information and they have other roles. The important thing, it's the observation that led me to think about something like Gong at the time, is that what they're not doing. And CRM was built for the leadership team. Before that, our salespeople were using Rolodexes or simple software do, they could manage the accounts, but we wanted something that's central that we could run our reports like every Monday and see what kind of revenue we should be expecting. And we started chasing them to keep the system up-to-date. It doesn't help them in a fundamental way. They could run their 20 contacts in the Google sheet. That's good enough. So it's not the easiest thing to use. I mean, it's not super-hard, but it's a chore, then I need to go and do my homework, fill in the data, the data as we've discussed, isn't necessarily very accurate in representing the reality out there. And it doesn't do a whole lot for me as a sales person, it's good for my managers. Gong solve these two problems. First, it doesn't ask the salespeople for anything. You just talk. And that's it. The system already, think of it like a fly-on-the-wall that just captures everything and updates the system. And everybody needs to know what they need to know without anybody having to lift a finger. And second, it gives them some pretty useful insights. If anything, just to kind of refresh my notes from my prior... If I'm meeting with a customer, I can go back to listen to my previous conversation, just so I'm up to speed. It gives us some metrics, things like the Fitbit for conversations. How can I be more interactive, ask better questions, provide better experience to my counterpart.
Brian Kenny: Alison, have you looked at CRM systems, maybe in the research for this case or in other contexts?
Alison Wood Brooks: It's funny. The shift from CRM, which has been so dominant in so many firms for so long, to this sort of new, this revolution, as we're shifting to companies like Gong, mirrors a shift that I've had much more contact with myself in behavioral science, moving from sort of traditional social psychology methods, like self-report. When we want to know if people are anxious, we ask them, "Brian, how anxious are you?" And you have to spend time telling me how anxious you are on survey items or something. There's all kinds of problems with this, exactly like Amit was describing. There's subjectivity in it, there are problems we don't remember. If I ask you how anxious did you feel yesterday? You probably don't remember very accurately. And so, there's been this shift, this massive shift, a monumental shift from self-report and observing big decisions like, are you going to marry this person? Are you going to ask for a higher salary? There's been a shift to just capturing people's behavior and especially their conversational behavior so that we're not missing out on that richness and we're not requiring people to go through the effort of self-report. It feels like there's a revolution afoot. And the same is true in the sales domain here, we're moving from CRM to methods like Gong that are a lighter lift and provide just so much more useful data than self-report.
Brian Kenny: So we've heard a lot about AI. I think some people are a little afraid of it and what the implications are, because we've heard some doomsayers talk about AI and robotics. And this seems like a fairly benign application of AI, but I'm wondering, maybe you could explain Alison, how does Gong work and what kinds of insights can the AI actually provide?
Alison Wood Brooks: Yes, so AI is not my bread and butter, but here, let me tell you what I know about it. At least in the context of conversation, people should not feel threatened by conversational AI yet, and they shouldn't for a while. There is so much nuance and so much we have to learn about how human beings converse with each other. To me, there's this really big question too, for conversational AI development, which is when we're designing chatbots, are we aiming for them to be like human beings with flaws and all, because that's what makes a lot of our conversations the most fun, the most interesting, the most challenging, the most uncertain, the serendipity of it, when you can uncover things you didn't expect to learn from each other. Or are we trying to design them to be some sort of perfect version of humanity that outshines humans and doesn't make mistakes? Both of those things by the way, are decades away from actually happening, which is surprising, because if you see little videos online from things like Google assistant. There was this video that went viral a couple years ago of this chatbot calling maitre d's at restaurants and making reservations. And it's very impressive.
Chatbots can do really impressive things. They can, especially in very constricted domains, like making reservations, they can understand a difficult foreign accent, they can recover from disfluencies in the conversation. They can ask repair questions like, "Oh, I didn't hear what you said. Can you say that again?" So they can do amazing things, but it's in very narrow domains for very specific purposes. And there's so much about human conversation that relies on our individual personalities and personal traits and our relationship like with each of you, I've now known Amit for almost five years, and we've had really interesting touch points of working together. Brian, same with you. And it's important when we converse with each other, that I'm able to go back to our shared experiences and shared reality and both feel that we have this common ground together and be able to raise it in useful ways when we gather again and chatbots really can't do that yet. So anyway, that was a long and windy way of saying-
Brian Kenny: I feel better though. So thank you. Thank you-
Alison Wood Brooks: Yes, humans are great. And I really love thinking about where we are in developing conversational AI and where it's going and where it could go and how can we use it usefully?
Brian Kenny: So you set out to solve some very specific kinds of issues Amit, when you came up with this idea, how is it going? What kind of insights are you gathering from how people are applying Gong?
Amit Bendov: It's very exciting. First, we have in a short amount of time, almost 2,000 companies now use the product, tens of thousands of people. The most surprising thing is the emotional reaction that people love. Just people absolutely love the experience in the product. It's something that they haven't thought about. And you said there's some initial concerns, "Oh, these people will know what I'm doing," but after the first day or two people are so excited and we measure a net promoter score. And Gong almost always scores in the high sixties and low seventies. For reference, this is higher than the iPhone in 2008, which is very surprising for essentially enterprise software, usually not the most exciting stuff, kind of the people absolutely love the experience and those new capabilities. So that was a big surprise and it's very exciting.
Alison Wood Brooks: The cold call of the class is, does recording our conversations help or hurt? And often before people try Gong, they're really afraid. They're really afraid of being recorded, that it's going to be disruptive to the flow of their conversation, that they're going to be distracted, worrying that who's judging them and who's going to listen to them. In practice, and the experience of being recorded in a conversation, you very quickly forget. Because conversation is this really cognitively demanding environment that requires us to have this sort of reciprocally sustained involvement with another human mind. So many other things just melt into the background when we have this sustained involvement with another human mind and the same is true when people use Gong. And then they get this amazing, these beaucoup benefits at the end, like Amit was saying, you can listen back to yourself, you get all of these metrics about how you're doing and how you can do better. And those benefits outweigh, in experience, the scariness of being recorded.
Brian Kenny: So Gong actually, it doesn't just allow the sales person to hear the recording. They're providing more value add than that. They're doing some coaching they're trying to help them learn and gain insights, because I think, we're all inundated with data. Everybody's got lots of data at their fingertips. The question is always, what does the data tell me? How do I act on it? Gong has sort of addressed that. Can you talk a little bit about what they're doing on that front?
Alison Wood Brooks: So, there's this enormous richness of data. And then, okay, what do I do with this? Do I just listen back? I don't have time to listen back to myself. I don't have time to analyze the data. So Gong beautifully sort of does that all you. They analyze dynamics like talk ratio, what percentage of a sales call was the salesperson talking versus the customer? The general recommendation there is that the sales person doesn't talk for any more than 65 percent of the time. And actually, closer to 50-50 is better. They measure the longest monologue. So in duration, in time, what was the longest speech made by a single speaker on the call? The longest story, this is the duration in minutes of the longest customer talking segment. Interactivity, and this is a score to show how often the conversation switches back and forth from team member to customer. So this is an indicator of turn-taking, which is a really important aspect of conversation. It's how dialogue is different than one-way communication like public speaking, is that we take turns, we can go back and forth. And so, this bubbling level of interactivity is a signal that people are into it. They're succeeding in this sustained involvement and they're doing the bubbling discourse that defines good conversation. And Gong also provides this patience score, how long a salesperson waits after the customer has stopped talking before they take over the conversation. I like this, because what's sort of hard is that you want to score high on interactivity, but you also don't want to be interrupting them. People don't like being interrupted. So you want to be both high in interactivity and high in patience and that's a hard balance to strike.
Brian Kenny: This has been a fabulous conversation. I've got just a couple more questions before we go on. I will say that I was a little nervous engaging in this, because I feel like you're probably dissecting everything I'm saying, you're hearing me this very-
Alison Wood Brooks: You know what's so funny Brian? I get this a lot, because of my research and teaching. But one of the biggest takeaways from the whole course and all of my research is that you can learn tactics, you can learn strategies, you can learn how to prepare better for conversation, but it's not a checklist. The thing that's beautiful about conversation is that it's serendipitous and unpredictable and you can't control or predict exactly what your partner is going to say. So once you're in it, you kind of got to let it all go. And I really don't sit and analyze and dissect everything because man, would I be miserable if I did that?
Brian Kenny: You'd be exhausted.
Alison Wood Brooks: Exhausted and miserable, and you'd miss out on the joy and delight and really the serendipity of getting to interact with other human minds. And that's what I love and continue to love.
Brian Kenny: So let me just ask you, you mentioned AI before and your thoughts on AI, do you see this movement as something that's going to continue to grow? I mean, Gong, there's a couple of competitors in the marketplace, do you see this as kind of the next wave of CRM?
Alison Wood Brooks: I do. I do. Gong has a few competitors in the market, in the sales domain. There are lots of other emerging players outside of the sales domain as well. So we do research partnerships with people across all kinds of domains, not just sales. And so, there are companies like Gong emerging in education, in work meeting management. Anywhere where there are people having lots of conversations that could be recorded and analyzed these types of analytics can be helpful. Sales is a logical one to be a leader, because they have objective outcomes. They can measure, how did I talk on the sales call? How did it influence whether that customer wants to talk to me again? Did that call convert to a sale? What do they think of me? They have very clear objectives, which is not as clear in other domains, but still these technologies are emerging in other domains as well. And I do think it's the future. In terms of AI and chatbot development, every large company is working so hard on chatbot development. It's sort of this undeniable thing, more and more data is available to us. What are we going to do with it? Oh, we're going to use it for many different purposes in many different ways. I'm excited to see where it goes, but also wary of the ethical implications of what's coming our way.
Brian Kenny: Yes. So Alison, one more question before we go. And that would be, if you want our listeners to take away one big idea from this case, what would that be?
Alison Wood Brooks: Conversations matter. We hear a lot about the art of conversation, but there's really more science than art in conversations, and now we can measure it and analyze it and do it better.
Brian Kenny: Alison Wood Brooks and Amit Bendov, thank you so much for joining me today. It's been great having you on the show.
Alison Wood Brooks: Thanks so much for having us Brian.
Amit Bendov: Thank you very much.
Brian Kenny: If you enjoy Cold Call, you should check out our other podcast from Harvard Business School, including After Hours, Skydeck, and Managing the Future of Work. Find them on Apple Podcasts or wherever you listen. Thanks again for joining us. I'm your host, Brian Kenny, and you've been listening to Cold Call, an official podcast of Harvard Business School, brought to you by the HBR Presents network.