Derek Smith - SVP, Innovation & Data R&D at ZoomInfo

(Joel Beasley at 00:00:00) Hello, my friends. Today we are talking to Derek Smith, the CTO at ZoomInfo, and we discuss where the line is drawn between AI algorithms and queries, the right and wrong ways to give feedback, and how to prioritize the product roadmap to drive innovation. All of this right here, right now on the Modern CTO podcast. Here we go. This is the Modern CTO podcast.

(Joel Beasley at 00:00:33) Hello. Hello.

(Derek Smith at 00:00:35) Hey, Joel. How's it going?

(Joel Beasley at 00:00:37) Amazing. Let me just get the audio through my AirPods here. Boom. All right. Can you hear me all right?

(Derek Smith at 00:00:44) Yes. Loud and clear.

(Joel Beasley at 00:00:46) Nice. Dude, I'm a big fan. I'm a big fan of ZoomInfo.

(Derek Smith at 00:00:50) Me or ZoomInfo?

(Joel Beasley at 00:00:51) Not you.

(Derek Smith at 00:00:51) Not you. Okay. Good. Yeah.

(Joel Beasley at 00:00:53) We'll decide at the end of the conversation. No, but we switched to ZoomInfo, I think three months ago, and we were going through this whole thing of, you know, who do we have on? I was like, let's go look at some of the products that we like and go talk to those people. And ZoomInfo, it really helped our sales team with the data, and we just re-engineered our sales process. And your data was a part of that, and it's been working out really well since. So we had an efficiency improvement through being able to do automation, less manual collection time spent, which just means, you know, more outreach, more dollars, more salespeople, and then you can grow the company and apply that culture and have a great life.

(Derek Smith at 00:01:40) Yeah. I mean, we exist because sales can be a really inefficient process. And what kind of salesperson wants to spend their time researching data or guessing emails? You know, that's just not an efficient way to use someone's time who's really skilled at selling. I mean, that's an entry-level job, and they're spending a lot of their time doing that. And at ZoomInfo, we just give that data to people to make them more efficient and more successful.

(Joel Beasley at 00:02:04) How did you meet these people? Like, how did you meet the people at ZoomInfo and get to that position?

(Derek Smith at 00:02:09) Oh, this is a good story. I was in law school, and the guy seated next to me was starting a company and wanted some help. Henry Shuck, the CEO, decided to start a data company, software company, in the middle of law school. And so I helped him out with some crawling and, you know, menial data work. And actually, I have a huge passion in sports. So I left the company in 2009—and the company at the time was called DiscoverOrg, and there's been lots of mergers and acquisitions that have created—I think we'll talk about it later because it's interesting from a leadership perspective. But I went to work in sports, and I kept my eye on what Henry was doing in Vancouver, Washington. And the company kept growing and growing and growing. And in 2016, he basically said, you know, it's now or never if you want to come back because we're taking off, and I don't have time to keep asking you to come work for me. So I was tired of New York City. I went out there, and it's just been a whirlwind ever since, just growing. And now I'm on the leadership team and working on one of the fastest-growing software companies out there. So that's how I met up with all those people.

(Joel Beasley at 00:03:27) That's amazing. Did you play sports when you were younger?

(Derek Smith at 00:03:30) I played baseball. Baseball is my passion. I actually worked for Major League Baseball for six years in New York. And, but I realized at about age 18, I was not going to be a Major League Baseball player. And if you can't play Major League Baseball, well, why don't you try to work in Major League Baseball? Right? At least you get a taste of it. And I feel very lucky to have fulfilled basically a version of my childhood dream of being in Major League Baseball.

(Joel Beasley at 00:03:57) Right? That sounds so cool. What did you do there? Like, did you get to work on—what type of technology—what are you doing at—what does technology people do at MLB?

(Derek Smith at 00:04:06) So I was doing investigations. So if you think about steroid investigations or fraud investigations, there is a lot of data wrangling that goes into cross-referencing, you know, player—let's say a player tests positive for steroids. Well, he's not going to tell you where he got the steroids from, but we want to find out where he got the steroids from, right? Because maybe other players are getting them. And so you have to look at who the player leaves tickets for, who they trained with in the off-season, where they used to live, who do they live with in spring training. And all this data is in different places, and you have to bring it together to try to find out, like, who else is doing steroids and who's the doctor. And that was a big part of my job. And, you know, ultimately, at Major League Baseball, I was finding information in different ways and then sometimes contact information for Dominican prospects. So there's this problem in the Dominican Republic where people will lie about their age, right, because documentation isn't great. And so someone tries out when they're 14 years old and they're not very good. They come back four years later, and they still say they're 14 years old. And they look big and strong and they're amazing. And, you know, unless you have systems to capture biometric information or to document, you know, what they look like or where they were born, you're going to get defrauded. And baseball teams are going to spend millions of dollars investing in this player that they think has lots of room for growth that actually doesn't.

(Joel Beasley at 00:05:34) That's amazing. So I worked on this project, like, 10 years ago. And these guys would come in around lunchtime and it was a side project of the organization I was working with. And they were building and buying teams down where you mentioned, and they gave that exact thing. They were saying, oh, you should—they were showing pictures of these 14-year-olds that looked like they were 25. And I was like, there's no way. And it's so competitive. I don't understand the business model too much. I just know you can start a team and somehow that becomes profitable, because I didn't get into it. I was just like, oh, here comes, you know, the group of guys that has the baseball team down in that other country. And it was just unbelievable. But that is exactly how it works, and that's so true.

(Derek Smith at 00:06:22) Yeah. And the other thing that I worked in in Major League Baseball that I think is fascinating from a management standpoint is salary arbitration. So most people aren't familiar with how this works. But there are a set of players in Major League Baseball that at the end of the year, they have to negotiate their salary with their employer, the team. And if they can't reach an agreement, they each submit which number they think is appropriate to a panel. And they argue for their numbers to the panel, and the panel has to pick one of the numbers. They can't pick a number in between. And even more complicating the process, the panel doesn't know anything about baseball. And so you have players saying, I should be paid a lot of money. I'm as good as these other guys who they're probably not as good as. And then you have the team management saying, you're not that good. You made all these mistakes last year. We think you should only make this much money. And then the arbitrators pick one of the numbers, and then you have to go back to work with the other party the next day. So, you know, it's a system designed to encourage settlement between employers and employees. You know? Who wants to go through that process? No one. Okay. Let's meet in the middle. But sometimes you don't meet in the middle, and you have these court hearings where baseball players are basically arguing with their teams about how good they are. And that's a very data-driven exercise where you're trying to project how much should I pay this guy based on the statistics? You know, how much can we afford to pay? Who is he similar to? Who are the most similar players? And, you know, it's the fusion of statistical analysis, the law, and sports. And that's, like, basically my entire background rolled into, you know, one two-hour hearing.

(Joel Beasley at 00:08:03) That's amazing. What was the culture like? You know, I would imagine it was pretty competitive, but did that carry over from the sports world into the MLB organization? Was it a competitive work culture?

(Derek Smith at 00:08:15) You know, I worked for the league office. So we were kind of like a shared services organization. So we were the overarching entity, you know, organizing baseball. But—well, sometimes when you're doing investigations in steroids, you feel like the police. Right? You know, you're—but I would say that the competition between teams is so intense. Right? You are—you know, Moneyball. Right? Moneyball is a really famous way that Billy Beane and the Oakland A's use statistics to get an edge. And everything's an edge now—training, nutrition, travel. Like, you know, you're trying to just turn the screw a half screw tighter to get that edge on the competition and to find that inefficiency that no one else has found. And so, I mean, one of the problems is when you have that competitive of an environment, people do break the rules. You know? Like, oh, you're not supposed to steal other people's coaches. Well, who's to say you don't send a text message to the guy, you know, and tell him how much you want to pay him even though it's tampering? And then how do you prove that they tampered? So it's really hard to enforce rules when the stakes are so high.

(Joel Beasley at 00:09:24) Yeah. It sounds like they have a bunch of small rules. Like, I didn't know about any of those. When you can't steal your coach, it's like, that's not free market. Like, if I want to pay someone to do something, it's their decision. Right?

(Derek Smith at 00:09:37) Yeah. Well, if he's under contract, you know, there's long-term contracts in sports. And the benefit for the coaches, if you get fired, you still get paid for the next four years, multimillion dollars. So, yeah. It's, you know, sports and business, very different, but I was very lucky to find some common threads that have really helped me in my role today.

(Joel Beasley at 00:09:58) Yeah. No. That sounds unbelievable. I was reading your article about AI, and that made me a fan. You had an incredible amount of unique thoughts about it. But can you give me an overview of what GPT-3 is?

(Derek Smith at 00:10:16) Sure. So GPT-3 is basically—I don't even know how to explain it, but it's a language service that can basically replicate or tries to replicate human responses, right, and the way that humans use language. And it also has this entirely massive dataset behind it of everything that it's crawled on the internet. So it knows how certain things are worded, how certain responses are given. So you can do things like, say, you know, tell me a story about baseball written in the tone of Emily Dickinson. And it will do something like that, you know, and it will attempt to answer any question possible. So you can say, you know, who's the president of the United States? And it'll say Donald Trump. You can say, you know, who's the host of the Modern CTO podcast? And it'll say Joel Beasley.

(Joel Beasley at 00:11:05) That's right. Amazon Alexa will tell you that too.

(Derek Smith at 00:11:09) Oh, really?

(Joel Beasley at 00:11:09) I didn't know. Yeah. Yeah. I said, who is Joel Beasley the other day? And she spit it back. And I was like, whoa.

(Derek Smith at 00:11:16) That's pretty cool. I think I'm not high-profile enough to get that, and the Smith last name makes it pretty unlikely that I—

(Joel Beasley at 00:11:24) Derek Smith. That's like you got to try even harder now. That's what you have to do.

(Derek Smith at 00:11:28) It's good when people try to Google me though if I want to be hidden. You know, I'm harder to find.

(Joel Beasley at 00:11:33) That is true. Until those AIs, until that GPT-3 gets all over you.

(Derek Smith at 00:11:37) Yeah.

(Joel Beasley at 00:11:37) Tracks you down.

(Derek Smith at 00:11:38) And, you know, GPT-3, people are lauding it as, like, you know, the—like, this is the human—this is the takeover. Right? AI is taking over for humans, and there's articles in favor of how amazing it's going to be. There's other more critical opinions of it, but it really does make it very easy for someone to leverage machine learning, natural language processing, and really intelligent types of computation that wasn't available before.

(Joel Beasley at 00:12:06) Yeah. I've read this column on, like, the Guardian, and it was entirely written by GPT-3. Have you seen that?

(Derek Smith at 00:12:16) No.

(Joel Beasley at 00:12:16) Okay. So they said that it was written by GPT-3. And the purpose of the article, like, the motivation they gave it was to write an article and convince humans not to be afraid of AI. Right? And then they said that they took some editing. But as I read it, I was like, there's no way. Like, the way that it was written, I was like, there's no way that GPT-3 did this. So my production team and I, we start digging deeper. Okay. So apparently, they produced a series of articles, and then they cut out the text that makes the most sense from all. And I said, that's like me following my three-year-old around with the camera for six months and then stitching together some insightful clips to make her sound like a savant even though she's just—you just edited it. Like, that doesn't count. Has to be one article and it has to be fluid.

(Derek Smith at 00:13:05) Yeah. In my personal opinion, it does some really cool things, but it doesn't do anything well enough to be relied upon all the time. And that's pretty much what we need in technology and in business is something that's reliable. Right? And if you ask it how many eyes a spider has, it'll tell you eight. And that's because the eight legs and spiders and eight arms, it's always going to have these things that it gets wrong. And that's why I think that, you know, the fusion of human intelligence and machine learning and intelligence is really where you can make the most impact in business. I don't think—obviously, all manual work isn't going to work these days, but also, we're—at least we're very far, I think, from relying fully on something like GPT-3.

(Joel Beasley at 00:13:59) Yeah. I mean, well, I'm excited about Neuralink too. But when I see those, what gets me excited is watching the rate of progress between the years. I think it was like 175 billion points. I had like more than doubled from the previous year or two years, but I just fast-forward it in my head 10 years. Right. At least there's—something's happening. People are doing it and it's useful. But, you know, I don't really get scared about the AI stuff. It doesn't really scare me at all because, worst-case scenario, we can just turn off all the computers. Right?

(Derek Smith at 00:14:38) Yeah.

(Joel Beasley at 00:14:38) Once they have bodies, that becomes a little bit more difficult. Right? Like, once we have more humanoid shells, like, if there was an autonomous factory generating these autonomous shells that were all connected, then you get issues because the machines could actually force you to physically do stuff. But we're not there yet. Right?

(Derek Smith at 00:14:56) Yeah.

(Joel Beasley at 00:14:57) So that's—we're not—I think Will Smith made a movie about that.

(Derek Smith at 00:15:00) I think there's a lot of movies about that.

(Joel Beasley at 00:15:03) Like Roomba would be the one to blame. Right? If you want to see the seeds of this right now, go check out a Roomba. Okay. So as an engineer, I want to ask an interesting question.

(Joel Beasley at 00:15:17) Where do you draw the line? Or I guess let's make it more of a discussion. I was trying to figure out, where do you draw the line between, like, AI queries, algorithms? And I was wondering, have you come across, like, a nice infographic or some nice consolidated explanation of the differences between these things? Because it's just so ambiguous, AI.

(Derek Smith at 00:15:38) It's very ambiguous. We did an internal podcast for our whole sales team, you know, because we sell a product that's heavily based in AI, and they're talking to prospects every day about what we're doing in AI, but they don't even know what AI is. Right? Or they don't know how to explain it. And so I think there's—I found, like, a Venn diagram. It kind of said, like, this is AI. Inside it is machine learning and natural language processing and deep learning. And then data science kind of overlaps part of the AI circle, but there's other parts of data science that aren't in AI. You know, like EDA or, you know, who knows what else. And so AI is data science, but there's other data science that's outside of—and, like, algorithms, you know, that are probably outside of AI. And then there's stuff within AI. But I do think the definition on a macro level of AI, it kind of changes over time. At least this is my—you know, when I think about artificial intelligence, it's like it's a process or a machine that's doing things that require human thought or human decision making. And before there were calculators or abacuses, like, you know, humans had to count in their head, and then we built machines to do that. So maybe, like, you know, a thousand years ago, an abacus was artificial intelligence, you know, or a calculator is.

(Derek Smith at 00:17:05) And so I think as machines get more sophisticated, the things that humans do change over time. Right? And as what humans do over time—now the goalposts are moving, and now you need machines to do what humans are doing today. And then in twenty years, the goal is going to be to have machines doing what humans are doing in twenty years. And so I think artificial intelligence is almost just on this journey where it's lagging behind the tools it makes. And I view it just as human-like decision making and computational abilities and, you know, broadly applying those principles to what we do.

(Joel Beasley at 00:17:52) Have you written an article on that? No. That's a really good thought. Like, when you're explaining it, it was unique for me. I had never heard it explained like that with the abacus and we're always trailing, and it is really interesting. And it's also interesting how, you know, I read in your other article, the companies, the startups that had AI were getting, like, 50% more funding, and it's interesting because it's like these buzzwords. It's like as a collection, if we look at ourselves as, like, one organism, it's like we want these things to sort of exist. And so the money part of our organism just throws money at these certain—at these buzzwords. It's like as much as people knock on buzzwords, their importance is, like, critical. Like, they're just absolutely required. It's the only way we can communicate, like, cross-industries and get new products to market quickly. You know, I used to look down on the fact that a new technology or something would come out, and then the board would tell the company, oh, we need to implement this type of—you know, we need AI. We need AI in our company. You know? And it's like, well, it's because that's the future, and you need to be thinking about the future.

(Derek Smith at 00:19:03) Yeah. No, I think there are people jumping on the AI bandwagon, though. You know? I think it's also become a marketing device. Right? Like, oh, like, I'm looking at buying this golf club, and it was designed with AI. And, like, I don't know what that means. You know? Like, I mean, you can—they're like—you just put AI next to a product, and all of a sudden people think it's fancier and it's smarter. And, like, I think in my article, I pointed out that, like, if you have .ai as your domain, you get three and a half times as much funding. Now, is that because you have the domain or is it because you're just doing something that's the future? You know, we can argue about that. But there was a study that said 40% of companies that claim to use AI aren't actually really using AI. But there's no AI police. Right? There's no one going around saying, like, prove to me you're using machine learning.

(Joel Beasley at 00:19:56) No. That's my new job.

(Derek Smith at 00:19:58) Yeah. I believe. It'd be kind of fun. It could be a TV show that I'd like to watch that I don't think most people would, but it's like, why investigate if you're actually legitimately using artificial intelligence? Because you get away with it. No one's going to say, show me the proof that you're using AI. It's just—no one gets to that level. The proof is in whether your product is compelling. And compelling products aren't really going to have a long shelf life if they aren't using these types of techniques because that's what you need these days to stay in the game.

(Joel Beasley at 00:20:31) Yeah. You need, like, you string together, like, all the buzzwords, like quantum computing, AI, Neuralink interface. That's what we're building here.

(Derek Smith at 00:20:40) Yeah.

(Joel Beasley at 00:20:40) And the checks just start coming in.

(Derek Smith at 00:20:43) And no one on, like, a sales call is gonna be like, wait a second. Can you explain to me, like, what those—like, no one in today's world, people are kind of embarrassed to, like, say, can you explain that to me? Right? Or what exactly do you mean here? And so the conversation just keeps going because no one wants to get into a detailed discussion of what, you know, Neuralink is or neural nets are on a sales call.

(Joel Beasley at 00:21:07) Yeah. What did you—you mentioned an acronym earlier about data. You said, like, EDA? I had never—

(Derek Smith at 00:21:13) Exploratory data analysis. So, you know, if you wanna build an algorithm or a model or predict something, like, you don't just—you don't just start on day one crunching numbers and trying different combinations of features. Right? Like, you need to understand the data. Right? Understanding, you know, like, garbage in, garbage out. Right? If you put dirty data into a model, you're just gonna get dirty data out. So this is part of my job at ZoomInfo: building the algorithms that do things like predict the employee count, that predict the revenue for a private company, that predict the likelihood that this person is still employed at this company. To do that really well, you need to understand what data is available to you, how it got there, how it changes over time, and the best way to use it. And, you know, that requires some research before you dig into the real data science. And it's just exploratory data analysis. It's digging into—it's basically evaluating the ingredients that you have before you bake a cake. You know?

(Joel Beasley at 00:22:16) So you guys are using, like, legit AI in the product?

(Derek Smith at 00:22:20) We are. Yeah. I think, uh, you know, DiscoverOrg, what the company used to be, was 100% manual research at a time. So we were just a CRM of data that was manually gathered by humans. And so it's kind of funny. In 2016, our platform at DiscoverOrg only had 800,000 contacts. Today, we have 120 million. And so that just shows you—and in 2016, we were the leader. You know, we are the leader. So in the past four years, the cloud and big data has just totally, you know, flipped the script on what you need to do to keep up in a data-focused sales environment. But what we have the ability to do now at ZoomInfo is we're not having the researchers call up and say, hey, what's Joel's phone number? You know, when you have fifty, seventy million phone numbers, one more doesn't really move the needle. But if you have those 150 researchers augment your algorithms, find out the inefficiencies, you know, give that ground truth feedback back to the data scientists to improve the algorithm, all of a sudden, you're optimizing algorithms like no one else can because none of our competitors have a research team of 150 people augmenting AI algorithms. So—

(Joel Beasley at 00:23:36) Well, you built your business model, like, to include that, right, from the get-go. And it's gonna be hard for anyone who has a different business model to provide that same competitive advantage with their current cost structures. Right?

(Derek Smith at 00:23:49) A hundred percent. And, you know, business data is hard to keep up on. So, you know, like, we're not—you know, we're a software company that—we're also, could say, we're a data company. But, like, you know, selling personal data, it's pretty static. I have the same cell phone number I had twenty years ago. I have the same personal email I'll probably have for the rest of my life. Like that data, when people sell it, it's still good. But business data changes all the time. I mean, I've had three different employers. I've had three different emails at this company. I've had seven different direct dial phone numbers in my three different employers. And that data changes all the time for people, for companies. You know, there's mergers and acquisitions, companies grow, companies change technologies. BB&T acquired SunTrust last year or earlier this year. Their company size went from 15,000 to 35,000. They moved headquarters. They have a new CEO. And someone needs to keep track of all this to help people know who to sell to, you know, to prioritize their accounts. And that's what we do. We try to keep our arms around everything that's happening in the business world so that sales and marketing teams can effectively find the companies and the people that they need to sell to.

(Joel Beasley at 00:25:04) Have you guys ever bought, like, an email marketing company or, like, a reply? I know I use reply.io, but you ever buy a company like that?

(Derek Smith at 00:25:14) We did. Um, so we acquired a company called Tellwise two years ago, which is, basically—I think we're now calling these, at least on the sales side, it's called sales engagement. And it's basically you can now pre-schedule outreach. You know? So I think a lot of the listeners here that are in technology, they probably get these emails, you know, asking them to, hey, are you interested? Will we take a call? Most of those emails today are, like, totally pre-planned. You know, they're obviously not—

(Joel Beasley at 00:25:42) Are you sure?

(Derek Smith at 00:25:43) Yeah. Oh, isn't someone writing three paragraphs just for you, Joel? Um, you know, they're recycling content. And they're not clicking send. Right? They are scheduling, okay, I'm gonna reach out to Joel. If he doesn't answer a week later, I wanna send him another one. If he doesn't answer that, I'm gonna schedule a call in because I have his phone number from ZoomInfo. And then, you know, on average, it takes, like, six to seven touches before someone actually says, yeah, okay, I'll listen to what you have to sell. I think it might help us. And how you optimize that outreach is really what you were kind of talking about earlier: how can you be efficient in going to market in selling your product? And it's really by automating a lot of the sales and marketing messages that are sent.

(Joel Beasley at 00:26:32) Yeah. Because I was thinking I was like, I wonder if ZoomInfo would ever, like, build a connection back to, like, Reply because all this—you—we take the data from you and then we go do stuff with it, but you don't get that information reported back to you. And I was like, I wonder if they're gonna buy one of these companies because it would—it would help, in my imagination, which is up for debate. It seems like it would help to have a feedback loop on that. And if you owned a company that was also sending the messages, you would have a greater detail of feedback.

(Derek Smith at 00:27:04) Yeah. So—feedback—that's a great topic, feedback loops. We actually use feedback loops a lot in keeping our data fresh. So, you know, when customers integrate their marketing automation system with ZoomInfo, one of our conditions is, hey, we'll help enrich your data, but we want you to share with us when an email is delivered and when it bounces. Right? We're not going to look at the content of your email. That's none of your business. We're not going to look at who you're sending to. You know, that relationship is private. But telling me that, you know, derek.smith@zoominfo is valid or bounced, that really helps us keep our data up to date. You know, we have an accuracy score in our platform that lets customers know the likelihood that this person is still employed. And so if I see that Joel Beasley opened an email yesterday, he's probably still employed. Right? So we can take that and our algorithm will increase the confidence that Joel is still employed. If Joel all of a sudden starts bouncing, that's a problem. You know, something happened to Joel. Um, maybe you just rebranded. Maybe who knows? But the feedback loops are a big part of what we do at ZoomInfo.

(Joel Beasley at 00:28:13) That's why you gotta get—like, that's why you guys probably bought, like, NeverBounce. Right? I actually—I use that tool independently of you guys. And then the other day, I was on the tool, and I saw at the bottom that you guys own them. I was like, what? That's crazy.

(Derek Smith at 00:28:27) Yeah. It's, um, you know, what happened was, in 2016 or 2017, like, whether an email is valid or not is a great way to find out if someone still works somewhere. Right? And I don't wanna get into email deliverability too much, but, the spam filters are getting more complex. They're getting smarter. Gmail now has the tabs for promotions versus inbox. And if your company's domain rep—and I'm sure a lot of the IT guys that listen—if your company's domain reputation is bad, your life is not fun as someone in technology. Right? Because your sales and marketing team, they're not getting responses to their emails. You know, we had an issue, like, three, four years ago where our DocuSigns weren't getting into the inbox, so you couldn't even get deals signed. And so sending emails to bouncing emails is a really bad thing to do as a company.

(Joel Beasley at 00:29:18) It's, like, the worst thing you can do.

(Derek Smith at 00:29:20) Yeah. It's—and so how do you make sure that you don't send to bouncing emails? Well, you use a tool like NeverBounce, and it'll catch any of the servers that, you know, actually provide a response.

(Joel Beasley at 00:29:33) Yeah. Because I actually dug deeper into it, and it's like there's an array of responses. So you can sort of choose—we use, like, Zapier, and then I use filters based on the NeverBounce response. And I tested it with a couple different accounts just to see, like, you know, what filters matter the most. Uh, so there's always—it's always fun playing around and learning with that stuff. And I say playing around, but I mean, it's directly tied to revenue. It's, like, top of funnel stuff. It's how you get meetings and then how you get proposals and sales. Right?

(Derek Smith at 00:30:03) Yeah. And you know those feedback loops that we talked about, they feed the NeverBounce algorithm now. So if I see Joel Beasley sending an email yesterday in Marketo or in one of our community, now if someone else tries to check Joel Beasley's email in NeverBounce, it's gonna be valid because we have a lot of evidence that Joel is employed even if his email is one of those ones that when you ping the server, it doesn't give you a response. So, um, it was a very strategic acquisition for us because we make NeverBounce better, right, with the feedback loop data, and then NeverBounce makes us better by letting us know when a lot of those emails are no longer valid.

(Joel Beasley at 00:30:42) You guys just had your IPO. Right? You guys are on the Nasdaq now. It's big time. How was it going through that process? Did it affect you at all?

(Derek Smith at 00:30:51) Um, it definitely affected me. Yeah. I was, uh, so I was part of the roadshow team. You know? So, you know, we have these plans of flying on private jets from San Francisco to New York to Chicago presenting, and we were actually supposed to IPO, like, the third week of March.

(Derek Smith at 00:31:06) That was the day. And then obviously COVID hit, and we hit pause on that for about two to three months. And so we did virtual meetings, meeting with investors, lots of Zoom meetings.

(Derek Smith at 00:31:20) And, you know, it was an amazing event. It was a great part to be a part of. Would it have been nice to get pictures in Times Square, you know, and celebrate with my coworkers? Yes. But ultimately, we accomplished something that we had our eyes on for a long time that was really rewarding.

(Derek Smith at 00:31:41) So, you know, hopefully we'll do a one-year reunion or something if New York City opens up. But, you know, I never, when I was in law school and this guy asked me to work for his company that he's starting, never in a million years did I expect it to have a market cap of $15 billion and be on the Nasdaq in 2020.

(Joel Beasley at 00:32:01) You guys should have rung the bell virtually. I saw them do it. I was talking with Harry, the CIO of Zoom. They did it from space. They let the International Space Station ring the bell.

(Derek Smith at 00:32:12) Yeah. I think we probably missed out on that one. But, you know, we had a bell, and Henry had a bell in his office. So I think he probably did his own bell ringing, personal bell ringing.

(Joel Beasley at 00:32:23) Oh, that's a good time to celebrate. To be honest with you, when I saw you guys come up on the market, in my Robinhood, I bought some of your stock because I buy stuff for the long term, like decades. And I use the product, and I know the product, and I'd seen it around for years and years before we actually, you know, used it. And so I was like, alright, this is a company that I think will do well in the long run.

(Derek Smith at 00:32:49) Yeah. No, I'm confident in our team here at ZoomInfo and what we're doing. And, you know, one of the problems we have, and I'm guessing most leaders have this, is prioritization. There's so much we could possibly do. I think everyone probably feels under-resourced. I need more people. We have all these great ideas. What are we going to do to impact the business the most? And they're hard decisions.

(Derek Smith at 00:33:04) You're just making calculated bets on which products are going to provide the most value to customers, where the market's going. And I like where we sit, but, you know, I'm sure Nokia liked their position ten years ago or whatever it is, right?

(Joel Beasley at 00:33:31) So how do you do that? Like, how do you decide how to place bets? Do you have a system or a process? Do you have a new product team that explores and launches and tries to gain traction? Like, how do you do that as an organization?

(Derek Smith at 00:33:45) Yeah. So that's actually kind of close to what my job is now. So I'm the SVP of Innovation and Data R&D. And what we struggled with at DiscoverOrg and ZoomInfo for a long time was we were spending so much time combining our companies. So DiscoverOrg was really the combination of two companies that were the same size. They were both about 300 people, and we both did the same thing. And so you spend all this time combining two companies that do the same thing to get more revenue, right? Huge boost in inorganic revenue. And then you do it again for ZoomInfo, two companies of the same size, and we did the same thing very differently.

(Derek Smith at 00:34:30) ZoomInfo is about big data crawling. Get as much data as you can. DiscoverOrg's all about accuracy, right? Highest percentage of deliverable emails. And so you had two different philosophies, two different religions, and you had to fuse them together really fast. And as I'm sure most people can imagine, the systems you need to collect as much data as possible are very different from the systems you need to collect data as accurately as possible. And so we were just spinning wheels trying to get these two systems to work together to make our existing product as good as possible. And no one was thinking about, well, what are we going to sell a year from now? What are we going to sell two years from now, right? Like, we can't just be this dinosaur that just, you know, has emails and phone numbers, right? We need to move forward. So I kind of gave up all the data and the research team, and now I'm focused on innovation.

(Derek Smith at 00:35:25) And the way I do it, at least for the projects that I'm assigned, is it's a lot of research and it's a lot of conversations. So advertising. We don't do anything with advertising right now, but we probably should. You know, we have marketers as customers. If you want to send an email to 500 people, you probably want to place ads to them too, right? So why can't ZoomInfo be part of that process? And so I'm meeting with consultants. I'm talking to people. I'm building up wireframes. And I love it because I'm not really fixing problems anymore. I'm coming up with ideas.

(Derek Smith at 00:36:04) And would you rather come up with cool ideas or fix problems? You know, one of them is very stressful. The other one is really rewarding. And so I think I got the good end of that deal. But it's still hard. Even when you know what you want to do, it's like, okay, everyone agrees this is super important to do, but we have that other thing we said was super important four months ago and it's not done. And you just kind of stack it up and you do your best, I guess.

(Joel Beasley at 00:36:33) Yeah. It's hard not to get distracted by the flavor of the week type deal. Because, you know, you hear things like, oh, we get into marketing, or when I was thinking of future stuff for you guys that I would be interested in, I was thinking audiences, right? So you're familiar with audiences in marketing. So I think audiences, and then to get better than that would be if you acquired some company and had first-party intent data. So, like, I could, we could, that could be a filter, right? Like, when you're searching, it's like show me all the people that match these labels and has intent data that matches this. And now it's like, boom, I've got people that are in a buying cycle. You know? There's enough data out there to connect all those things. You just need a lot of money.

(Derek Smith at 00:37:19) Yeah. Yeah. And luckily, we kind of have, you know, as a public company, we have some cash to go acquire companies and to maybe put our foot on the gas pedal in terms of acquisition. You know, we were so worried about margins and valuation when we were a private company, right? And now I feel like we still care about those things, right? But we're bigger, and we have more infrastructure to handle acquisitions, to invest in new products. And I think the two areas you mentioned are definitely on the road map, but not in the long term.

(Joel Beasley at 00:37:57) Yeah. Having cash or, what's the popular word now? Having dry powder, right? I think that's what the VCs are saying. That's really, really useful to the growth of a company. And I'm still learning. Like, our company has about ten people and I've been, and, you know, we've been up and down and we're, we had our first two years where it was really, really difficult. Our third year, we figured out who we are and what we're doing and who buys and why and, you know, redefining all of our processes. But all of this experience of running a business from an operational level, like you said, caring about margins and cash flow. And just because you hit your sales goals doesn't mean you have enough cash that month. You know? It's just given me an entire insight, a whole new perspective, because my previous, like, ten, fifteen years was just building apps and selling them. So I'd get a contract and fulfill it. And so I said I own my own business, which I did, right? But it was a different thing than operating a B2B sales organization business. Very different.

(Derek Smith at 00:39:00) I kind of learned myself that there's a big difference between doing things and running things, you know, or managing people. And, you know, it's kind of the same and kind of different, but I've always been a data guy. I like digging into data. I like drawing conclusions. I like coming up with solutions. And when you start getting into management and managing more people, you start separating yourself from the groundwork, right? And it comes down to building teams, and it's a different skill set needed. It's a different type of thing. And I found that I kind of like managing smaller teams, you know, ten people, five people, more so than 200.

(Joel Beasley at 00:39:45) How is it different? Like, how would you say, because typically you would only have, like, five to seven direct reports. So is it the conversations that you have that you get to have with those with your direct team at ten plus people that are better than at 200? What's the difference?

(Derek Smith at 00:40:06) So, you know, I mean, I think everyone's different, obviously. But for me, the difference with a small team is that I'm 100% invested and interacting with 100% of my team. You know, everyone in my org has one-on-ones with me and knows that I'm looking at their projects and that I'm helping them. And we have this really good culture. And now all of a sudden when I'm managing a 200-person team, I don't have the ability to get to know all 200 people, right? I don't have the ability to be present and to build those relationships. And sometimes I would try too hard. So a good example is after the ZoomInfo acquisition, I knew that everyone on my team was in a good spot. You know, M&A, everyone worries about, like, oh, is someone going to take my job? Am I going to get replaced? And I knew everyone on the team was in a good spot, but not everyone knew that. You know, they didn't, everyone just kind of looks over their shoulder all the time, right? And so there was a lot of friction there because I didn't communicate as well as I should have at the time.

(Derek Smith at 00:41:14) And so I tried to overcorrect, and I scheduled one-on-ones for all 27 people on that data team over the course of a month. And I learned an amazing amount of things. You know, people will tell you things in a one-on-one that they'd never just volunteer to you. You know, someone told me they don't have anything to do in the day. Like, I'm not really working on anything during the day. And I'm like, we are so busy. How can you not have anything to do? And he didn't tell me that until we were getting coffee down the street. You know? And he never would have just knocked on my door and said, hey, Derek, I haven't done anything in three days. You know? And that's great to want to be a part of what everyone's doing, but that's not how you scale an organization, right? And so I think the difference is you have to empower and trust your direct reports. And I'm kind of a hands-on guy, and I held on to too much. And I didn't really, I guess, deed enough over to my direct reports.

(Joel Beasley at 00:42:13) Boom. Dude, that's one of the most common stories I hear.

(Derek Smith at 00:42:16) Yeah. And to be honest, I was listening to your podcast on my way to Montana because I was like, oh, I'm going to be on the podcast. I should find out, you know, what Joel talks about. And as I was listening to it, I was like, you know, I wish I had listened to this podcast a year and a half ago because it could have helped me go through that process. You know? And I listened to Roy from Calendly.

(Joel Beasley at 00:42:40) Oh, he's great. Yeah.

(Derek Smith at 00:42:42) And then The Trade Desk's CTO, you know, that was a good one. And I just started realizing it's like there was an example on one of your previous episodes where some guy went to NASA, and he was told, you're only going to be evaluated on how strong your direct reports are, or something like that. You know? And that's the thing I didn't do when I managed 200 people. You know, I did not view it as my job to actually build a bunch of VPs that can go do what I do. And that's what I 100% should have been doing. You know, build people who can do all the stuff I'm doing for me, can take it over so I can kind of just oversee it. But I was trying to do too much and it doesn't scale. And so that's my tip to people trying to grow into larger roles. You know, I went from managing four people at Major League Baseball to 200 at ZoomInfo in four months. And that's a big switch. And so anyone who's taking on a bigger team, find the VPs, find the people you can trust. If you don't have good VPs, develop the directors into VP level so that you can just get out of some stuff. Because if you don't have people under you you can trust to do a core part of your job, your organization's responsibility, then you're already, you're going to lose. You know what I mean? Like, you just have to get that level hammered up.

(Joel Beasley at 00:44:06) Yeah. I 100% agree. What I'm thinking of right now is when I was an engineering leader, I understood what it felt like to have someone I trust. Like, you can give something to them. They're always on time. They're always ahead of schedule. Everything just works. It's always yep and good news and a good attitude. And it's like, you know, I spent a lot of time building a really solid engineering team that I did many projects with for several years. But for some reason, it didn't click for me to do that with the sales part of the organization when I started a real company. Like, it's like, you know this lesson, but sometimes they don't transfer over and there's no one that's going to walk up and tap you on the shoulder and help you out. So I found that the best way to deal with these things is just by listening to people that are smarter than you, asking questions, doing some learning, but then also making sure that you take actual action because that's when you get the experience and it solidifies and you have a story to tell and you remember it for a long time.

(Derek Smith at 00:45:10) Yeah. You know, I know there's a common engineering saying, like, a tremendous developer is worth ten mediocre developers, you know, in terms of their output. I would say for VP or director level of management, it's like, you know, someone you can just give stuff to and forget about it. I mean, the mind space that that opens up for a senior executive is just, it's so coveted. You know?

(Joel Beasley at 00:45:39) When you're developing them or when you're talking about the concept of developing them, in my experience, you know, I'll find the right person that has the potential that you can, after you have success with one or two people, you can kind of find that type of person, right? It's like, I'm looking for these traits. I'm looking for these principles. But a lot of people talk and I think it's a little ambiguous about developing. They're like, oh yeah, we're developing VPs or developing these people. And I'm always curious to dig under the surface of that. And it's like, what are the actual tactical things that you do to develop your VPs?

(Derek Smith at 00:46:17) Yeah. So at ZoomInfo, we've really invested a lot in courses. You know, so leadership development there, I think once a week for eight weeks, you come and you learn about prioritization, having tough conversations. You know, I'm sure a lot of the stuff that, I mean, you could speak a lot better to than I could. But the way I do it is you kind of know what someone's missing, right? You know why this person's good. They have these skills. There's someone on my team who's the smartest guy, knows our industry inside and out, easy to work with, but didn't always speak up, didn't always assert himself in conversations, right?

(Derek Smith at 00:46:53) And never really had to have tough conversations. So for him, you know, I view it as kind of rounding someone out, right? That they do this stuff good, they need to build up these muscles to be the total package. Right? And so you just kind of, I'm a big fan of just example-based learning, you know, and it's like, hey, in that meeting, I could see you wanted to say something, you didn't. Like, you have to speak up every time in that situation.

(Derek Smith at 00:47:22) Or like, hey, I noticed this person's doing this. This is probably a good time to have a tough conversation with them about, you know, coming into work on time or, you know, making sure they work a full day or not being on their phone at their desk. Right? Even those little things, kind of like, you know, building those. I really see them as muscles. Right?

(Derek Smith at 00:47:41) You know, with muscles, the more you build them, the bigger they get and the easier they are to use. So I try to find the weaknesses in someone. And admittedly, I wasn't that good at this. Right? So maybe I'm not the best person, but that's how it's worked for me in the past is, you know, identifying the person and then finding those areas where you can make them better.

(Joel Beasley at 00:48:02) No, I think it's wonderfully said. Right? Because it's not gonna be the same for everybody. It's not like, you know, pushing people through a program. That's like one aspect of it. Right? That's like, here's some foundational knowledge so you at least know this information in the universe exists. You can go dive in deeper later if you need. But then that idea of like, and it goes back to you getting to know your team really well. The more time you have to spend with your team, the better you understand them as people. And then the more qualified you become to coach them through getting out of their comfort zone. And it's not like you have a checklist of a perfect employee. You just, the only way to really do it is to get to know them and then grow them. Push them out of their comfort zone.

(Derek Smith at 00:48:45) And I think you said something important there. You know, there's no such thing as a perfect employee. I think also giving your direct reports responsibilities that mesh with what they're capable of. You know? I mean, so if you have a great leader who understands data and everything, but they don't have that presence, you know, that public speaking skill, like, maybe don't have them run the all hands in front of 200 people. Right?

(Derek Smith at 00:49:09) And so, like, just putting people in a position to succeed and, you know, maybe carving out little things that aren't suited for them and giving them to someone else. You know, I think the work that you give someone, you can give people jobs that they're destined to fail at. And if you give someone a job that they're unlikely to be successful at, that's your fault. That's, you know, it's probably more your fault than their fault, right, in assigning the work.

(Joel Beasley at 00:49:37) Yeah. I know. It's back to your, you know, muscle reference. There's challenging and then there's, like, injury. Right?

(Joel Beasley at 00:49:45) You don't wanna handle that much weight. Right? If you see them and they're a hundred and ten pounds soaking wet, you're not gonna hand them like a fifty pound dumbbell. Right? They're just, it's, they're not gonna have success. They're gonna get hurt with it. But you have to kind of figure out where they're at and then give them weight that's like, you know, not too easy, but is also challenging so that they grow a little bit.

(Derek Smith at 00:50:05) Yeah. And professional psyches can be fragile. You know, that one comment from the CEO in a meeting, I mean, that can bring someone down for a month, you know, like an entry level employee. And so...

(Joel Beasley at 00:50:19) They're an experience though. Right? Like, I think I am very mentally tough.

(Derek Smith at 00:50:25) Yeah.

(Joel Beasley at 00:50:25) And so it's hard when people who aren't mentally tough meet me because they think I'm... Weaker people don't like the truth. They don't like it.

(Derek Smith at 00:50:37) I know. There's a phrase. One of my flaws is that when I think something isn't a good idea, the feedback that I give is a little too transparent. You know? It just kind of like, I cut to the... I'm truthful. I give you my opinion.

(Derek Smith at 00:50:55) And, you know, that's good in some ways, you know, because I'm not hiding the ball. I'm not dancing around the issue, but it just comes off wrong. And one of our executives says, I don't suffer fools gladly. I guess that's a phrase. You know? But yeah, I think people who do give direct feedback and are very upfront and candid, that is overwhelming for some people, you know, to give feedback that way.

(Joel Beasley at 00:51:23) I think about it like, I'll, like, on computer, it's like, alright, let me, here's the response, but now let me wrap it in a nicer human layer sometimes.

(Derek Smith at 00:51:32) I learned something from one of our executive coaches that I'll pass on that I... It's called LCR. It's like, concern, recommendation. So anytime someone tells you something that you don't think is a good idea, start by finding something in their proposal that you like. You know? So maybe you're like, hey, I think we should put purple bushes in front of our house.

(Derek Smith at 00:51:58) And then you go, oh, I think you're right. That's a good idea. We should put bushes in front of our house. One of my concerns is that the purple won't look good with our house color. I recommend using a lilac or a blue.

(Derek Smith at 00:52:13) And all of a sudden you've turned what I would probably say is like, that stupid purple would be ugly. You know? And you give them feedback and they kind of feel like you're on their team. Like, they like the bush idea. You know, they're with you and they're tweaking their idea.

(Derek Smith at 00:52:32) But my mind, I'm probably more similar to you. It's like, no, that's not a good idea. I just say that. And so ever since I heard LCR... Like, concern, oh, no, it's LCS, like concern, suggest. Ever since I got that, it's really been working.

(Joel Beasley at 00:52:49) Yeah. I had some self-reflection time when, after, like, the third time I was called flippant. And I was by these older people, and they didn't know each other. And it was just this time in my life where people are saying, you're being flippant. And I'm, so I Googled the word, and I was like, yeah, I respond quickly.

(Joel Beasley at 00:53:08) And so, you know, I've learned one of my favorite ones is from a Navy SEAL called HALT. So if you're hungry, angry, lazy, or tired, like, don't make decisions.

(Derek Smith at 00:53:20) Oh.

(Joel Beasley at 00:53:21) Don't make decisions if you're hungry, angry, lazy, or tired. So you make your worst decisions when you're in those states. And so that's one that I heard once. And for some reason I just can't get it out of my head because it's just so true. Like I notice myself whenever I'm, like, I'm trying to make this big change right now because I'm, you know, unhappy or whatever it may be or I need something different. And I'm always like, wait, I'm just tired. I'm exhausted. I need to go sleep and then consider this in the morning.

(Derek Smith at 00:53:49) Yeah. I think I should extend that to not just making decisions, but, like, don't react. Yeah. Because, like, you know, someone messages me something on Slack that upsets me, and it's happened a lot lately. And I get this mini rage.

(Derek Smith at 00:54:04) And it's because I'm, number one, like, I was here when this company was started. Like, this company is in my blood. You know, I'm so passionate about it. It really is my life right now, this company. And so I care about, you know, the location of a button in the platform. Like, it's the most important thing in the world, and I care about that button location at least fifteen times more than anyone else in the company. And so when I don't like where it's going, like, the passion in me just gets so strong. And sometimes I answer in a way that is just too flippant or too aggressive and, you know, toning down that intensity. And, you know, in the end, it's just the button. Right?

(Derek Smith at 00:54:46) And I don't always know the best place to put the button. My opinion is just one of many. And trying to scale that down to a place where my feedback can be constructive and not overwhelming or scary at times is something I'm really working on. And I'm getting a little better at it, but I certainly have more room for growth.

(Joel Beasley at 00:55:06) Yeah. Delaying responses is by far, like, one of the most useful professional skills I've ever come across. Like, just that extra twenty-four hours in that email, just coming back and looking at it when you're in a different state. Like, whenever I'm like, you know, you did that motion with your body, like, oh, I'm gonna respond to this person. And so I'm like, I catch it early and I'm just like, no, no, no. I'm just gonna go walk away and come back later when I'm clear.

(Derek Smith at 00:55:34) I think we just came up with a feature for, like, you know, Lenovo or Apple. If the intensity of the key presses is too heavy, it'll pop up like, you seem angry. Are you sure you wanna send this email? I would pay money for that right now.

(Joel Beasley at 00:55:53) Yeah. Yeah. That, or they're looking at your camera too. They could... I'm sure they're already doing that.

(Derek Smith at 00:55:59) Oh, yeah.

(Joel Beasley at 00:56:02) Dude, this is great. I really wanna respect your time here, your hard stop time. Was there anything else that you wanted to get out, or that we... First of all, there's twenty things that we didn't cover. So you have to come on in six months or twelve months, and we have to... We'll do, like, a two-hour podcast because I really like you, and I think we'd have a really good long conversation. But is there anything else that we didn't get out there in this episode?

(Derek Smith at 00:56:24) You know, we're just... I mean, ZoomInfo is a great place to work. We've got offices in Boston, in Tel Aviv, in Grand Rapids, Vancouver, Washington, Washington, Bethesda, Atlanta now. And so we need really talented engineers, smart people in technology. We have lots of jobs open. I'm hiring a data scientist in Portland. So, you know, if anyone's looking for a really good environment to do some really cool stuff, real-time streaming, big data, whatever it is, we're doing just about everything. You know? And we need talented people. So anyone out there who's looking for a change, we're glad to bring you on.

(Joel Beasley at 00:57:03) And where do I send them? Do I send them to your LinkedIn? Do I send them to ZoomInfo's career page?

(Derek Smith at 00:57:09) Yeah. Anyone can message me directly on LinkedIn, and I'm happy to pass them to the right people, to our recruitment team. But I think it's careers.zoominfo.com. But let me just...

(Joel Beasley at 00:57:18) Let me just check it out. We'll take a look at it, and we'll put it in the show notes too.

(Derek Smith at 00:57:22) That's perfect. Yeah. If ZoomInfo careers, just Google that. You'll find it.

(Joel Beasley at 00:57:26) Yeah. Dude, this is awesome. You have a good time. We made a podcast. How do you feel?

(Derek Smith at 00:57:30) I feel great. I love talking about this stuff. So whenever you wanna talk more, just let me know.

(Joel Beasley at 00:57:36) Dude, absolutely. You have a fantastic day, my friend.

(Derek Smith at 00:57:39) Okay. Thanks, Joel.

(Joel Beasley at 00:57:42) Thank you so much for listening. And if you found this episode useful, please share it with a friend or colleague who you think would get value from it. And if you have topics that you'd like to hear discussed on the podcast, either add me on LinkedIn or send me an email, [email protected]. Every time I get an email or LinkedIn message, it absolutely makes my day and inspires me to keep going.