Atlas EditorsPower to the product: Three founders reveal how they built industry-changing products customers love
In the last six years, the cumulative market capitalization of companies that prioritize product-led growth has multiplied more than 100x. It’s clear that an outstanding proprietary product is critical to winning—and keeping—customers. However, this is much easier said than done. The rate of startups that fail absolutely dwarfs those that are successful.
So what does it take to become one of the rare few who build industry-changing products that customers love? As part of this year’s Cloud 100, we sat down with three brilliant innovators who achieved staggering success to find out—Amit Bendov, Co-Founder and CEO of Gong.io; Neha Narkhede, Co-Founder and CTO of Confluent; and Ali Ghodsi, Co-Founder and CEO of Databricks.
Find an underserved market niche and listen intently
Amit Bendov, along with his co-founder Eilon Reshef, built an enterprise product with an NPS in the 70s—an almost unheard of feat in B2B technology. (To put it in perspective, that’s higher than the NPS for the iPhone in 2008.) How did he do it? He listened to a market segment that nobody else was listening to—and he empowered them to do their jobs better than they imagined possible.
Gong achieved this NPS by enabling all customer-facing folks at a company to uncover customer insights when others in the market exclusively catered to leadership. By democratizing insight-gathering, the team freed employees at the bottom of an org chart to surface challenges and opportunities independently—a result that leadership teams love.
“Gong is the first tool that caters to everyone customer-facing—the sales team, the customer success team, the product team,” says Amit. “They absolutely love Gong because it just does everything automatically. It taps into conversations and can mine information without having to fill in forms. People are addicted.” Amit attributes Gong’s explosive growth to focusing on a market segment others overlooked.
Fittingly for a product that “listens” to and analyzes conversations using AI, Amit is steadfast in his belief that listening is the most important leadership skill. “When we have internal debates, I say, ‘Don't listen to me, I won’t listen to you. Let’s go and talk to customers,’” he says. “We'll immediately record conversations with users. Then we can come back to the conversation armed with data.’”
“It doesn't have to be words,” he continues. “It could be numbers and research, but we always prioritize listening to customer needs—whether they're explicit or implicit.”
Listening to customers has underscored the need to make Gong a seamless and automated experience. “We’re pioneering an autonomous application,” says Amit. “Look at Google versus Yahoo. Yahoo was a human-curated system—people would list the top 12 pizza joints in San Francisco. Google said, ‘Forget it. We're going to index the world's information automatically.’ Gong is doing the same thing—we capture the information from the ether without anybody having to lift a finger,” he says.
“We believe this is the next generation of computing and applications, and we're just overwhelmed by possibilities.”
Radically rethink accepted concepts to create a new category
Neha is a founder who accomplished something countless entrepreneurs attempt but few achieve: She created a brand new category.
She achieved this by challenging accepted industry norms and radically rethinking them from the ground up. “When we started, we observed that existing data infrastructures just do not support a continuous real-time flow of data,” says Neha. “And we traced that back to a fundamental architectural flaw in how systems worked.”
“Creating something like Kafka actually wasn't the first step on that journey,” she continues. “The first step was to hack together existing systems to see if we could just make them work. And when we realized it was more of an architectural flaw, that’s when we took a step back and envisioned this world where the whole foundation of how data could move and be captured in a company would change.”
This epiphany—that data needed to be dynamic and not stagnant—formed the basis of Confluent. “Our mission is to empower organizations like Goldman Sachs, Intel, and Capital One, to put their data in motion and harness value from their data in real-time to create new products, new customer experiences,” she says.
But while defining a new category hinges heavily on a category-defining product, that alone is not enough, says Neha. “Capturing the love and the mind-share of the user—in our case developers—had also been a fundamental piece of creating this category and making it stick. We had to capture hearts through storytelling around the product. Not to mention that it’s a long, arduous 10-year journey.”
Today, Apache Kafka is used by 70% of the Fortune 500. But it’s only the beginning, says Neha. “Many are still understanding the impact of data in motion to their businesses. Technologies like Kafka, Spark, Confluent, and Databricks are really just a starting point. There's such an exciting path ahead of us to truly innovate the core infrastructure that companies build around.”
Make bets in product futures—and defend your convictions from naysayers
In 2021, Databricks achieved a $28 billion valuation—more than 10x higher than their Series E valuation in 2019. How did Ali and his team pull it off? “Short answer is luck,” he says with a modest chuckle. “But if you want to really go back, I think we took a few bets that were considered kind of ‘out there’ at the time.”
“We bet on the cloud, which wasn't obvious in 2012,” he recalls. “We bet on AI, which wasn't widely known at the time. People thought of AI as robotics and machine learning that wasn't broadly used. And we bet on open source. None of those bets were obvious.”
“At the time, most people told us, ‘That's not going to go anywhere,’” he says. But these bets turned out to be spot-on, and when these technologies became mainstream, the Databricks team already had such a strong lead that it was near impossible for competitors to catch up.
“I think that's the way you build a great business,” says Ali. “If you're doing what everybody thinks is the obvious right thing to do, then what prevents the really big guys that you're competing with to do the same? And then you can't grow.”
“We got a six or seven-year head start to build a cloud product, figure out ML, figure out open-source monetization at a time when everybody else was like, ‘Those are small markets. They don't matter.’”
“That’s the key to success: Taking some big, bold bets and hopefully being right about them,” he says. “That's where you’ve got to cross your fingers because none of it will fall into place without some element of luck.”
Transcript
Ali Ghodsi:
Hey, everyone. Super excited to be here and have a conversation with you, Neha and Amit. Thanks for being a part of Cloud 100. I'm really curious to hear more about your thoughts around strategy and innovation and the products that you all have helped build. But before we start, I'm curious, what book has had the biggest impact on your approach to leadership? Maybe we can start with you, Neha.
Amit Bendov:
Neha, do you want to start?
Neha Narkhede:
For me, it's Mindset, where Carol Dweck discusses the power of a growth mindset. And it really resonated with me as it aligned very closely with the values I grew up with, which helped me to take the leap to start Confluent and also actually to overcome my fear of the open waters to learn scuba diving, but really as a leader, it has helped me think about motivation and influence people to overcome obstacles and really keep growing.
Ali Ghodsi:
That's fascinating. I think that's also Satya Nadella's his favorite book, and he made everybody at Microsoft read it as well. That's awesome, I really like that book, I have to say. And Amit, what about you? Do you have a favorite book that really influenced your leadership style?
Amit Bendov:
It's tough, there's a ton. I read hundreds of books earlier in my career, but if I have to pick one, it's probably The Warren Buffett Way. It's not [crosstalk 00:01:23] this particular book, there are quite a few, but I think it was eye-opening for me at first, if you can run an empire with just like five people, like Berkshire Hathaway, just some small company, shows the power of delegation and being like sound structures and culture underneath. And there's many insights like, "don't do anything you don't want to read about in the headlines tomorrow", right? It just... And how you run a business, I think it's incredible, super influential for me.
Ali Ghodsi:
That's fascinating. Actually, I'm curious. So you scaled lots of different businesses to unicorn status, and you've talked a lot about the importance of listening and that affected your leadership style. Can you tell us a little bit about that, and how that affects product strategy?
Amit Bendov:
Yeah. It is important to listen. A common mistake that people do is assume that they know, right? And I think humility is key to innovation. Just come with an open mind, don't assume that you know, don't come up with premature ideas. It's listening to customers, listening to sometimes what's not being said into implicit needs versus explicit desires. Go and talk with customers. So then we can come back with data, right?
Ali Ghodsi:
That's interesting. So at Gong, you actually listen to customer conversations, right? How does Gong work a little bit? Can you tell us a little bit about that? I find it super fascinating.
Amit Bendov:
Yeah, sure. So Gong is a system to really understand what's happening with customer, with customers or users, right? It taps into conversation, could be emails, zoom calls, phone calls, or any channel. You use your natural language, understanding to mine insights at a scale. So if you want to know, your competitors, people will talk about that feature or some other features or the new product launches successfully. Before Gong, you'd have to go interview salespeople or some customer, then you'd get like an input at best. Now you can know almost like in a second day, if a product launch is being successful and how customers feel about your product.
Ali Ghodsi:
I see. And are you using AI to do this? Or how do you actually listen to the conversations?
Amit Bendov:
Well, it is AI. So to understand what questions people are asking, what's their sentiment about a certain product or a certain competitor, what's working with the sales process? How is our customer's success experience reflect the way the customer responded? So it does understand quite a bit. It almost like magic. Obviously you can do the basic stuff, like listen to keywords. So product names, competitor names, but it can be higher-order things like, what are the action items? What's the engagement like on conversations? In a way that's very easy to understand.
Ali Ghodsi:
That's awesome. I wonder what it would say if it listened to this conversation. So the data is actually super important for that then, I guess. I'm curious, you know, I mean, you started a company that's focused on data, right? And data movement and data in motion? Could you tell us a little bit about Confluent and how it started?
Neha Narkhede:
Yeah, absolutely. At Confluent, our mission is to empower organizations like all the way from Goldman Sachs, Intel, Capital One, to put their data in motion and harness value from their data in real time to create new products, new customer experiences. And we started this journey by creating a really popular open-source project called Apache Kafka, and then later transitioning to providing both a cloud product and an enterprise software offering to essentially help organizations unlock their data and put that in motion.
Ali Ghodsi:
And with Kafka and now lots of competing sort of efforts around that, you basically created the category. Did you know that you were creating a category when you were at LinkedIn or is that that just something happened? Other people who are starting companies and they want to create categories just like you did, what can they learn from you?
Neha Narkhede:
Yeah, I think the reality is when we started, we observed that existing data infrastructures just do not support a continuous real-time flow of data. And we traced that back to a fundamental architectural flaw in how systems worked and actually creating something like Kafka wasn't really the first step on that journey, the first step was to hack together existing systems to see if we could just make them work. And when we realized it was more of an architectural flaw, that is when we took a step back and really envisioned this world where the whole foundation of how data would move and be captured in a company would change. And re-envisioned that as essentially Kafka that could fundamentally rethink of data as something that moves rather than something that rests and then occasionally moves around. And I think, fundamental to category creation has been capturing the love and the mind-share of the user in our case developers and that, along with creating a category-defining product, I think goes hand in hand in really doing that and not to mention that it is a 10-year journey.
Ali Ghodsi:
So in some sense, just focus on your user persona, the developers and the problem you wanted to solve, you know, there wasn't anything good out there and just solved that really, really well. And you then happened to also create a category along the way.
Neha Narkhede:
And then captured the heart of telling the story, which is equally important in category creation.
Ali Ghodsi:
So, Amit, I'm curious, you know, in the CRM space, there is like this 800-pound gorilla, right? Salesforce, super successful company, been around like, you know, huge. How do you carve out space for yourself among companies in that space?
Amit Bendov:
We created a new space, it does not replace CRM. Gong's the first application to focus on the customer-facing people, and they absolutely love it because it just does everything automatically, Gong just taps into the conversation and then it can mine information from there without people having to fill in forms. And it gives them insights, how they can be better with customers. And our net promoter score is almost always in the 70s, which is insane for an enterprise product. It's a consumer-grade NPS. It's higher than the iPhone in 2008. So by focusing on that persona, in addition to leadership, that really helped us drive a new growth. It's a new kind of system that doesn't take away from anything else.
Ali Ghodsi:
That's fascinating. And I have a question, so, okay. So you have awesome products and you know, you focused on the problem and the end-user and telling that story, but we also have to sell the software. So I'm curious, you know, since Atlassian came on the market in the early 2000s, I couldn't hear the end of it that, "Hey, they didn't need any salespeople." You know, it was product-led growth, the product led everything and you don't have to rely so much on with market and sales. And I'm curious about that interplay and what you guys think about it. So maybe, Neha, you can go first.
Neha Narkhede:
Yeah, I think with our cloud product and community downloads, Confluent has a very strong self-serve adoption model. Many prospects are in their earlier adoption stage before we ever engage the sales team, but a product that lends itself naturally to self-serve adoption is a good fit for building a product-led growth business, you know, open-source software has been a good example of that, as we've seen from the success of Spark Databricks and Kafka Confluent, but also another notable one has been self-serve freemium offerings that allow you to swipe a credit card and just experience the product, like Snowflake.
Neha Narkhede:
But I think what's less talked about is the art of layering on a sales sort of touchpoint and enterprise go-to-market motion on top of the groundswelling, right, that a product-led go-to-market motion creates. And I think that's a critical step on the path to building a multi-billion dollar company, which is really not only to build a bottoms-up adoption model, but to layer on the right involvement of customer success and enterprise sales to really maximize the value from the offering.
Ali Ghodsi:
Yeah. That makes a lot of sense. And I'm curious, you know, Gong must have an opinion on this because maybe we should all just be using Gong and we kind of automate away a lot of that sales process.
Amit Bendov:
Well, I think product-led growth is fantastic. Not every product lends itself to this, just the thing is they know how complex it is to understand, right? Our goal isn't to replace salespeople, but just take away all the stuff that they hate doing, like filling in forms, updating their managers, committing their forecast and all this stuff that nobody loves, right? But they're forced to do today.
Ali Ghodsi:
But certainly the cost of sales and marketing must reduce if you use Gong, no?
Amit Bendov:
Absolutely. First, like with product like growth and with Gong you could probably cut it by half. You take away like half of what people do, which is like not interacting with customers, it's just the hateful stuff. That's a huge saving and helps with ramp up time, increase quotas. The nice thing is it's easy to solve. Those are some of the stress points of that profession that maybe like notorious, there are movies about that cutthroat environment, all of that is not really necessary, but you could have like... You could make this fun.
Ali Ghodsi:
That's awesome. Databricks was the same, for what it's worth. When we started, we had Apache Spark open-source project and it had just gone viral and there was so many downloads, millions of people worldwide. And, you know, we started selling to small, medium enterprises or small and medium businesses mid-market but eventually to really move up and get those much bigger deals for the company and get the revenue.
Ali Ghodsi:
And you're growing, you know, significant, double, triple-digit, millions, you kind of have to have an enterprise Salesforce. That's what we learned because enterprises buy in a certain way. They have a procurement department, lots of people need to approve the budget. So eventually it turned out that really, to unlock that in the big enterprise, we have to layer on sales and marketing, but you still need a great product.
Neha Narkhede:
Couldn't agree more.
Amit Bendov:
A hundred percent. I got to ask you, if you don't mind. 28 billion? I mean, that's insane. It's almost like over a hundred X, right? From when you start? Teach me, like how did you do that? What drove that innovation? I want to know.
Ali Ghodsi:
Well, I guess... Short answer is luck, but if you want to really go back, I think we took a few bets and those bets were kind of out there, right? I think you all did the same. That's how you created the categories. And we bet on the cloud, wasn't obvious in 2012, we bet on AI, AI wasn't known at the time, people thought of AI as robotics and machine learning wasn't widely used. And we bet on open source, and those bets weren't obvious. And most people told us, "That's not going to go anywhere." And it turned out that those were the right bets. And eventually the market came there.
Ali Ghodsi:
I think that's the way you build a great business. If you're doing what everybody thinks is obvious and is the right thing to do, then what prevents the really big guys that you're competing with to do the same? And then you can't grow. So we kind of got 5, 6, 7 years to build a cloud product, figure out ML, figure out open source monetization when everybody else was like, "Those are small markets. They don't matter." And by the time we were kind of ready and the market came there, we had that headstart. So it wasn't that easy to catch up for the really big companies. So, I think taking some bold bets and hopefully being right about them and that's where some element of luck comes in?
Amit Bendov:
Yeah. You're being modest. This isn't luck, this is smart, like incredible, incredible, inspiring story.
Neha Narkhede:
I agree. So... Forgive me, sorry about that. On a completely different topic on all our ways of working have been upended by the events of 2020, but some businesses have navigated the transition really successfully, right? Well, I'm curious to learn from you guys, Amit, to begin with you, what technologies or tools or cultural transition did you have to put in place to ensure that a consistent spirit of product innovation was promoted while working remotely?
Amit Bendov:
Well, I mean, that's a great question. Obviously like a lot of companies, Zoom and Slack, would just communicate, communicate, communicate. So we want to make sure that everybody's connected. It was stressful and at the beginning and still, there're parts that aren't fun. I mean, we're doing very well in the business and I'm sure most of you are, but it was stressful, you know, it's just crazy times where like we have new hires out of college that not only didn't meet anybody, never worked at an office, don't even know what an office looks like, right? Maybe like from TV, right?
Amit Bendov:
So we used Zoom and Slack to communicate frequently with everybody, we used this little thing, Donut, right, that just helps people get a random coffee, which is great. And from a product... Obviously, we drink our own beer. So we use Gong a lot to communicate both with customers and with users, which really kept everybody connected. So people can immediately if someone has a challenging conversation and they're in Florida, right? Someone from California within like three hours can chime in and help and win as a team together. So that helped really help keep everybody together.
Neha Narkhede:
Yeah. I completely agree. Not just Teams, but we also learned that enterprise sales can be done over Zoom as well, which is something that nobody ever thought was ever possible. But now we all know we could cut down on some travel budgets.
Amit Bendov:
You know.
Ali Ghodsi:
I think good hygiene is important, right?
Amit Bendov:
Ali, what about Databricks?
Ali Ghodsi:
Yeah. I think good hygiene... Like those are things we should all be doing, but now we have to do them well, as you said, Zoom, Slack, you know, but I think meeting hygiene, which means make sure that the meetings have a good agenda, make sure you have good summaries, make sure that you have good written communication in the company so that people that are on different iPhones or at home, or couldn't attend the meeting can read up on stuff and it's clear. So I think those are things that everyone should do anyway, but I think it becomes more necessary as you go in this work-from-home environment. But I'll tell you this, this is our first week back in the office here. And I step out and I talk to a lot of the engineers, who just, you know, as you said, for the first time in their life, they're coming into an office.
Ali Ghodsi:
It's pretty amazing, I have to say, it's... I forgot how powerful it is to just stop by someone's desk and say, "Hey, who are you? Oh, awesome. Welcome to the company. What do you work on?" And they're super excited to tell you, I think that was really missing in the last year and a half. So I look forward to that, but I do think it's going to be a hybrid world where we're going to have some work-from-home, some in the office. And we frankly don't know, we'll learn as we go into this next sort of episode.
Amit Bendov:
Yeah. Neha, what do you think?
Neha Narkhede:
You know, I think we're entering in a hybrid world and I really am excited about that. But while some part of getting access to global talent is really exciting, being face to face and really connect with them is really essential. And I think this hybrid world is the new path forward.
Ali Ghodsi:
Yeah, it's, you know, and I think you could see the energy also from them as well. So I think this is something that I think we're going to relearn how important that was. And I think there's been... We probably over-rotated a little bit, people thought, "I don't need an office." And people, I can hire anyone anywhere and this is going to be amazing. But I think there are also a lot of advantages to the office so it's going to be some kind of hybrid that we learn. Also, so, you know, this has been an awesome conversation. I just want to sort of wrap up and ask you each. If you look ahead, when it comes to product innovation, what technology topic do you think more people should know about that would further your company's mission? Amit, maybe you can start.
Amit Bendov:
Well, we're looking at a little more abstractly, we're, we're pioneering an autonomous application, right? If you look at like Google versus Yahoo, Yahoo was a human-curated system, like people would put like the top 12 pizza joints in San Francisco and Google said, "Forget it. We're going to index the world's information automatically." Something's a hundred times more powerful and nobody needs to do anything, right? Gong is doing the same thing over like CRM. It's a new system that is autonomous application that uses AI to kind of capture the information from the ether without anybody having to lift the finger. There are other applications, right, that are actually more like Yahoo's human-curated system in healthcare, in HR, in other areas. We're just overwhelmed by possibilities, and we believe this is the next generation of computing and applications.
Ali Ghodsi:
That's awesome. Neha, what about you?
Neha Narkhede:
Yeah, I think, no surprise there, it is data in motion, at least two companies here focus on different parts of that equation. But yeah, Apache Kafka is used by 70% of the Fortune 500, but many are still understanding the impact of data in motion to their businesses. So there's such a large opportunity ahead of all of us here because organizations that can powerfully harness their data, whether it's analytical, contextual, intelligent, moment-by-moment will be the ones that create innovations of the future. So I really think that technologies like Kafka, Spark, Confluent, Databricks, and many more are really just a starting point and there's such an exciting path ahead of us to truly innovate the core infrastructure that companies build around.
Ali Ghodsi:
That's awesome. Yeah. And you know, I feel, for me, if it was something I would say, data and AI and combining data with artificial intelligence, machine learning and enterprises, that way they can sort of build the future and disrupt whole industries that they're in. I think I'm excited about that. And I think people are only realizing that we're just scratching the surface of that and it's going to have a huge, massive impact on everything around us. So I'm excited about that. So this has been an awesome conversation. Thank you so much for all joining us here at the Cloud 100.
Neha Narkhede:
Thank you. Thank you, Ali, for leading us.
Amit Bendov:
It's a real pleasure, I had a blast.
Neha Narkhede:
Likewise.
Ali Ghodsi:
Thank you.