Archives

Podcast Episodes

283. The Future of Cloud, Business Model Transitions from Subscription to Consumption, and The Shift from Technology-first to End-User Value (Dharmesh Thakker)

283. The Future of Cloud, Business Model Transitions from Subscription to Consumption, and The Shift from Technology-first to End-User Value (Dharmesh Thakker)


Dharmesh Thakker of Battery Ventures joins Nick to discuss The Future of Cloud, Business Model Transitions from Subscription to Consumption, and The Shift from Technology-first to End-User Value. In this episode, we cover:

  • Walk us through your background and path to VC.
  • What’s the thesis at Battery?
  • Any significant differences in types of products being built across geographies?
  • How does Battery segment the cloud infrastructure market and find opportunity areas in the subsegments of most interest?
  • To what extent do legacy IT and legacy software impede the rate at which we can progress with new toolsets and new infrastructure?
  • Is there more appetite for modern solutions, just due to demographics?
  • When ROI is opaque, how do you measure time to value and how might time to value might be different for different decision-makers within the organization?
  • What does top-down enterprise selling moving to bottom-up user and influencer adoption mean to ROI of significant scale implementation to new infrastructure?
  • What sorts of businesses lend themselves well to product-led growth?
  • You’ve said that managing churn and focusing on customer success is more important now more than ever. In what ways have you observed leading tech companies apply innovative approaches to customer success (expansion and retention)?
  • Should Startups price on a consumption basis early on?
  • Do you believe should the customer success leader report up through the CRL?
  • What are your thoughts on open source and licensing models? There are impassioned, differing viewpoints on this with some worried about the integrity of open source while others cast blame on large tech companies, like Amazon, that are efficiently able to monetize R&D that they didn’t invest in. What’s your position and how does it inform the way you approach investing in open source?
  • You’ve been investing in cloud since, I believe, 2008… what stands out to you as you look across the biggest winners? Any common threads or key differentiators?
  • What do you know, you need to get better at?

Battery Ventures provides investment advisory services solely to privately offered funds and neither solicits nor makes its services available to the public or other advisory clients. Nothing herein should be construed as investment advice. This podcast mentions certain Battery portfolio companies; for a full list of all Battery investments and exits, please click here. Content obtained from third-party sources, although believed to be reliable, has not been independently verified as to its accuracy or completeness and cannot be guaranteed.

Guest Links:

Transcribed with AI:

0:18
Dharmesh Thacker joins us today from Menlo Park. He’s general partner at Battery Ventures, a global technology focused investment firm. Prior to Battery, he was managing director at Intel Capital, where he led the firm’s global cloud and big data practice and made early and late stage investments. Dharmesh. Welcome to the show.

0:35
Nick, it’s great to be here. Thank you so much for inviting me. I’m a big fan of The Full Ratchet. And it’s great to be with you here talking one on one

0:43
Absolutely great to you know, reconnect and chat. So my my little intro doesn’t do it justice. But can you talk a bit about your background and your path to VC?

0:53
Absolutely. So you know, I’m an engineer by training, I got my undergrad in electrical engineering with chip design focus, believe it or not, so, back in the late 90s, I spent some summers at Intel working on Pentium two and designing chips in fabs and very quickly realized that I enjoyed interacting with humans more than I did interacting with chips and wafers. So post graduation, I spent most of my operating career in product management and entrepreneurship. There are a couple of startups, you know, join a public company has some early success selling my startups to other companies and security and cloud infrastructure and impose business school, I was looking to do my third startup, and we had our first child, and it became a tough choice, you know, do do another startup, which is this thing, having a full time baby or take care of the baby at home. And so, you know, I’m still married, which means I made the choice of not doing a startup and taking the next best thing I could do, which is join venture capital, you know, it’s, it’s amazingly stimulating intellectually, and it’s great to be involved with so many passionate entrepreneurs and, and be able to give back learning from my experiences and as an entrepreneur. And so that’s what I’ve been doing the last 10 years as a VC at first at an early stage venture fund called Advanced Technology Ventures, then Intel Capital, and then joined Battery about six years ago. But all that time spent in the area, I think I know, well, which is IT infrastructure and cloud software.

2:27
Very good. And can you I think most of the listeners know Battery, but can you give us that sort of the overview of the thesis?

2:35
Absolutely. So you know, most of you know Battery, we’ve been around for 35 plus years, manage roughly 9 billion assets under management, we do not invest in batteries, even though every every week, I get an intro from somebody from Asia Pacific on LinkedIn asking me about a new battery formulation. So we started on Battery Street in Boston, in the mid 80s. And one thing that has been true to our thesis and our strategy is refocused on doing a few things well. In our business, it’s about, we think it’s about doing one or two things really well and becoming subject matter experts and for Battery that b2b software, almost 90% of what we do is business to business software. And we try to find the best companies across the world, and really at any stage, but as long as the focus is business to business software. Half of it is on the infrastructure side, selling cloud data security software to an IT buyer. The other half is on the application software side where you might sell to a CMO, CFO, or chief revenue officer. But it’s all its offer. And we’re very excited that it feels like you know, software is still in the early innings. You know, technology as an industry and software only been around for a couple of decades. Whereas most of the companies that have been here for a couple of centuries have finally adopting it. So we’re pretty bullish about the adoption of software and the the market segment around software expanding from one to 5 trillion over time and still feel very excited about doing that. The one thing we have updated our strategy around is going global. You know, back in the 80s, you know, when a lot of tech entrepreneurship used to be in Boston then expanded to San Francisco. Over the last couple of decades, we’ve seen amazing entrepreneurs spring up over Israel over Europe, increasingly in India in the b2b software space. And so with that we’ve expanded our presence to to be where the entrepreneurs are. So our 10 general partners are spread across New York, Boston, San Francisco, London, Israel. And we can spend some time looking at the Indian ecosystem as well, trying to find the best companies wherever they are, and help them from the early stages.

4:39
Any significant differences in types of products being built across geographies?

4:45
For sure, but within the context of where we are, Nick, you know, our focus is very much on the US end market. So whether you started a security company in Israel or you started you know, a data middleware company and in Europe or us started a developer software company in India, you know, half the worldwide spend in these segments ends up in the US market. So we focus on these companies and help them move to the US as they’re scaling and expanding the US market. And that’s where we can help them with their go to market hires, customer connections, helping with you know, PR and team building. So regardless of where you started, our focus is only on those companies that they want to migrate to the US market, which we know very well, especially in b2b software. That’s not to say, there aren’t great companies that are being built, built for those local economies. So there’s great, you know, ride sharing companies and FinTech companies being built all around the globe. And we do some of that. But our b2b focuses very much on global companies that want to move and focus to the US market. In that context, there isn’t a whole lot of difference between them.

5:49
Got it. Got it. So Dharmesh, you talked about, you know, application software, you talk about infrastructure, as we think about cloud software at large. And the buyers, of course, you mentioned, you know, the buyers in the segment, they could be it versus a CRM, CMOS, etc. You know, how do you add Battery? How do you guys frame up? You know, the opportunities within cloud software and cloud infrastructure for that matter, you know, and then find the opportunity areas in the subsegments of most interest?

6:20
Absolutely. So, first I’ll clarify, Nick, you know, cloud is a fairly overloaded term, you know, these days and people say cloud, it kind of means software, right? So right, my focus within the cloud software segment is much more on the infrastructure side, where there’s an IT buyer. So whether it’s, you know, cloud infrastructure, developer tooling, data, and AI security companies, anywhere, there’s an IT buyer, either at an established company, or fast growing startup, that’s kind of my definition of cloud, right. And so, within that segment, you know, things are changing rapidly in the last decade, there’s a lot of kind of value creation in the picks and shovels, business, right, you have over a trillion dollars being spent in IT infrastructure on premises, which is moving to, you know, cloud infrastructure run by Amazon, Google, Azure, Alibaba, and others. And you know that that has created tons of value. If you look at the collector market cap of the cloud businesses that Amazon, Google, Azure, that totals over $2 trillion. And that’s based on just a 10% migration of legacy IT to cloud IT. So that segment, has done really well over the last decade, right, and there’s a lot of value increase. I look at the next decades now, the areas that I get most excited about are at the higher layers of the stack. And so the end goal of cloud IT is to free up companies from innovating at the application layer. So the first theory I’m really excited about within this cloud segment is just the software development lifecycle that enables rapid application development and innovation at a much faster pace than has ever been possible before, right. And that’s where companies like JFrog and Harness and Cypress that are in our portfolio as well as Git lab, hashey Corp and others that really speed up the developer cycle is super, super interesting, because ultimately, it enables innovation to move at a fast pace. And it benefits, you know, end customers. I think the second segment, which is also really interesting is the new innovation that we see, is powered by data. You know, if you look at like five years ago, most ecommerce companies or media companies that you look that had kind of like you know, an Oracle database or MySQL database in the backend with some data right? Now we have 100x, the amount of data from your social feeds to your location to your persona, there’s so many data sets out there that are powering applications in near real time. So you know, whether it’s you trying to find out every second where you Uber riders, or DoorDash order is, or whether you’re trying to predict how the COVID curve is going to flatten over time, there’s just a ton of data that is powering all these applications. So that I think is the second sub segment, the data economy powering innovation, which is super interesting. And then finally, just look at the look at how economical it’s become. Now if you look at the fact that cloud makes it so easy, data mining is so much easier, the end result is applications are able to launch at much faster pace, and at 1/10 to 1/20 of the cost of what was possible before. And as a result of that you just see much more innovation overall. So we look at you know, number of software companies that will create an operating a few years ago 10,000 companies, which has now expanded to 25 to 30,000 companies that are seed and venture backed right. So just the volume of innovation has become a lot higher thanks to the economies of scale that cloud can offer. And thanks to economic and data mining, and that I think is the most exciting aspect of my job. There are just so much to look at. It’s tiring sometimes because there’s a lot more companies and the signal to noise ratio is kind of low but at the at the margin, just the volume of innovation, that benefit customers and businesses is just unprecedented, which I think is the most, most exciting aspect of cloud and data the next decade.

10:07
You mentioned Oracle, SQL, you know, to what extent do does legacy IT and legacy software, whether it’s databases or cloud or on prem? Or, you know, what have you? How much does that impede the rate at which we can progress with, you know, a lot of new tool sets and new infrastructure? Because either learning how to communicate and play with legacy systems can be complicated, or, you know, just displacing them can be a challenge both technically and due to inertia.

10:39
No, you’re right. Look, I think, if you look at things pre COVID, Nick, you’re right, there was a slow rolling kind of transition from legacy IT to modern IT and cloud. Primarily, because data has gravity, you can just shut down a database and move, you know, the the core business data you have to a new system overnight, it’s a lot of hard work. So it takes time to make that transition. Which is why if you look at all estimates, you know, over the last decade, you now transition 10% of legacy infrastructure to cloud. And the expectation was that it’s going to take another couple of decades before the transition goes through. What happened though, is COVID, accelerated all of that, right. So you saw three to five years of transformation, from legacy to modern, it happened within five to 10 months, post-COVID, right. If you’re the CIO of a major enterprise, you want to maintain business continuity, and managing your data centers with remote employees, you know, resume sessions is no point, right? So that accelerated the condition for many legacy systems to the cloud providers. And you saw a major kind of disruption happened due to COVID. And some of those habits are not going to come back, right. So that dislocation happened much faster. And secondly, what you saw is, rather than lifting and shifting legacy apps over to the cloud, using companies, Reese rearchitected, refactor the applications using a new modern microservices architecture, because if you’re going to, if you’re going to use this downtime, and COVID, to to kind of transition infrastructure, you might as well go all the way to modernize the application, I generally would agree with you that transition would have taken a lot longer. But I think COVID has accelerated that. And I think you’re going to see a significantly faster transition path here. So that’s one aspect of it. The second aspect of this Nick is met new applications that are being built, a lot of the cloud adoption you’re seeing is from net new companies that didn’t exist five years ago. So if you’re starting from scratch, you’re going to start on the cloud, you’re going to start on open source on data. And so if you believe that, you know, the number of startups has gone from 10,000, to 25 30,000, in the last few years, most of those new startups are expanding the pie. And the new part of the pie is all in the cloud and modern IT right. So the collective impact of a faster transition. And net new companies starting on the cloud born in the cloud, is expanding the segment of modern IT at a much faster pace than we ever expected.

13:06
Also have to imagine that, you know, time goes on the the average age for CIOs and CSOs and whatnot, is is, is going down. And I think somebody I was speaking to just yesterday had mentioned that this is the last election cycle, we just went through, where the boomers are going to be the majority of the population in the US and next election cycle. It’ll be the millennials. And so I imagine, you know, there’s there’s more appetite for modern solutions, just due to demographics.

13:36
Yeah, certainly, right. I mean, you have a whole kind of whole generation that is used to instant gratification. You go to your iPhone, you swipe your photos, you have an aha moment, right? You sit on your Tesla, I mean, the entire experience is completely different. Right? You show up on Netflix, there’s a movie recommendation, this is the era of instant gratification. And are those guys come to work, running, you know, corporate systems and running, you know IT and DevOps, they expect the same kind of time to value from the software they buy, right? So there’s this been this whole generational shift where even the classic and I CIO buyers or security buyers, that used to focus on features and functionality and buy, like the most feature rich systems are now kind of seeing a trend shift to well, performing systems that deliver Quick Time to value right. Instant gratification is a way to land the deal, you can expand it with more features and functionality. But if you don’t deliver value in a clear way, you’re going to lose your your buyer, whether it’s a CIO or a developer. So I think there is a systemic shift to a newer younger buyer, who has much more emphasis on time to value because people pay for what they see. And if they don’t see value quickly enough, they lose patience and the more so I think there’s certainly a generational shift, which is impacting the way IT and software is purchased.

14:54
You know, I recently we had a discussion on the team here at New Stack and it was bit of a challenge, but it was we’re trying to unpack this time to value metric. And we recently did an investment in a company that helps reduce waste within grocery and restaurant environments waste food, by opening up access to data, actually just making sure that they know the inventory of various foods and explorations, and they can do much smarter ordering. And this company had one of the fastest payback periods and time to value of anything I’d ever seen. You know, within a month and a half of deployment, the customer had paid back the cost, which maybe suggests we should increase pricing. But nevertheless, my question for you, Dharmesh is, when you’re in an environment where ROI is a little more opaque, you know, efficiency gains, reduction in labor, you know, ability to increase top line by making better decisions, you know, on the revenue side with data. When it’s opaque, how do you measure time to value and time to value might be different for different decision makers within the organization? So you know, how do you think about that?

16:08
Yeah, that’s a good point. It’s a good point. And I would distinguish first between like, the time to aha moment, and then there’s a time to value right? what I was referring to earlier, instant gratification is like the time to aha moment, right? If you are, I don’t know, if you are like a developer that is, you know, building a new application, being able to launch a MongoDB Atlas cluster and run an application within 90 minutes and see it working. And that’s time to aha for that, right? If you’re like, you know, a cloud security architect, being able to see that 20 of your users are using sensitive data they should not be using, and that’s going to get you in trouble getting all of that within 90 minutes. So I think what we’ve seen, you know, we have this mantra of 90-90-90, within our kind of bottoms up, cloud companies where we say 90 seconds to like, download or get your Cloud account, going 90 minutes to sing the first aha moment, and 90 days to the first deal, right? So I was talking more about like the time, the aha moment. And that, to me is like perceived value, right? You see instant gratification, because you got immediate value, compared to the status quo, which would take you you know, months to deploy, let alone seeing the aha moment. And then once you see the aha moment, you’re like, 80% of the way there. The justification for ROI, in many ways is, is relatively easy in IT, right? Because if you’re a developer and you’re working Google, you’re probably costing, you know, seven figures a year. But let’s say a developer who cost the company 200,000 a year, right, being able to save, you know, 10% of the time and base, you know, kind of time being wasted on things that don’t want to do, that’s very easy to justify, if you’re selling me software that cost me $100 a month or 200 a month, 2000 a year to save me 10% of a developer costing me 200,000. The ROI is pretty straightforward, right? So I don’t worry as much about the time to ROI for IT systems because unlike groceries, where you may pay back in a month, but you’re talking about a good that costs you $10 in the case of IT or in the case of you know, data privacy or compliance, the stakes are so high. And if we can see the aha moment and you can see received value justification of hard dollar ROI is relatively easy, I think our most companies fail is they get so inundated by the complexity of deploying and showing all the features that you never get to the time to aha moment. So you never even have a chance to get to the ROI calculation. If you can show the aha moment, the ROI calculations are relatively easy and many segments of IT,

18:43
Okay, I guess that relates a lot to one of the major takeaways from your your open cloud, you know, report that you that you all at Battery he published late last year, but that there has been this massive shift, which is no surprise, right. But from top down enterprise selling to this bottom up user and influencer adoption, and if you can get tools and software and, and applications in the hands of users, and you get to these quick aha moments, then you don’t have to have this, you know, grand dog and pony show about ROI of, you know, significant scale implementation, you know, new infrastructure.

19:23
That’s right. Yeah, absolutely. I think you bring up a really good point. And again, this to me was one of the hardest things to unlearn and relearn, right? When I started my career in venture, the industry was all focused on building, you know, really sophisticated systems. And then you know, having an experienced sales reps, go sell it to CIOs, right and the, the community was around like the top 1000 companies that have more than a billion dollars in revenue. And you’d have the CIO or the CEO, CTO or the chief security officer for these companies. And you would spend six to nine months and sales cycles to these executives, and offer that close a seven figure to you, and then hope it would all get rolled out, you know, more often than not will end up on a shelf where because they never get adoption, but at least you made the sale. And you know, that’s kind of how business was done. And then starting 2015-16, it kind of felt like once Amazon had competition from Azure and GCP, you know, a lot of CIOs felt like they weren’t going to be locked into one cloud providers. So I think in general, they became a lot more open to adopting cloud. And once your users are used to operating on Amazon and Google Azure, they’re expected expectation was kind of what this over the iPhone like Quick Time to value, I can spin up an easy to cluster and Google Instant and BigQuery. And like, no time, right, so why is it that I buy the storage system that all of a sudden takes me six months, while my cloud provider is offering me the same functionality and features within 90 seconds? Right? So I think there was this seminal moment where buying patterns changed. And CIOs went from decision makers to influencers. And the practitioners, the end users are the ones that played a more important role in in the buying decisions. And if you look at the last five years, and look at most IT companies are more software companies that are not a 50 billion, right or not, the 20 billion even look at Snowflake look at you know, Data Bricks is private, but Twilio, Atlassian and MongoDB, you see a pretty important pattern that all these companies have nailed it. In terms of focusing on the end user, the practitioner, the person who is going to ultimately use the product, and directly go to them, sometimes using an open source approach, sometimes using a SAS approach, or both, and focus on time to value. And if you can focus on that the sales cycle is a lot faster, your land deals move a lot faster. And once you have your foot in the door, you can always expand from there and become an enterprise standard. Right? So it was kind of an unlearning and relearning moment for me. But that’s our focus going in most new companies, we focus relentlessly on the end user and focus on time to value for those end users. Because I think there’s a strong correlation between that and long term growth. Right. So that’s, that’s, that’s certainly an important aspect of this top down versus bottoms up setting. And the second aspect of this, which is interesting is, you know, many companies found out that the addressable market expanded by a significant factor, when you switch to the buyer will switch to the end user, right? So, for example, when you look at Atlassian, at the time of filing, I think they’re like 50,000 users, and now they have 170,000 users. And they say the addressable market is 1.5 million companies, right? So you wonder why is that the university is to be 1000, companies with more than a billion dollars in revenue that bought most of the IT systems, right. And now we look at many companies routinely reporting 20, 30, 40,000 customers, because all of a sudden, cloud made it economical for you to go sell to these mid market and SMB companies. And if you focused on the end user, you could focus on selling somebody $100 per seat license, and then expand to 10,000 users or 100 users over time, depending on the size of the company. So all of a sudden, by breaking down the barriers and going bottoms up selling not only to do you move the deals faster, but we also expanded the addressable market by a factor of five to 10x. And that’s what we see at play, the companies that when the bottoms up route, went after a much larger market, in a faster sales cycle, a more profitable sales motion, as a result of which many of these companies are at 500 to 700 million revenue run rates, still going 50-60% and much more profitable at scale, because they took their approach of focusing on the end user going through a product led motion, and focusing on a much larger market than was possible. So it’s a long way to answer your question, but you know, it’s an area we’re really passionate about. And this is kind of many of the criteria that we focus on new investments we make.

23:59
I love it. Competing with nonconsumption, right, find new markets and find new buyers and find new users as as long as you can drive the value you can you can capture some of it for yourself.

24:11
And shelfware That’s right.

24:14
So you mentioned MongoDB and we’re talking about sort of community driven platforms. You know, a lot of these are so good within the companies that this viral effect takes off this network effect and you know, devs share with other devs and users share with other users. You also mentioned the product lead growth motion, you know, where’s the the distinction between the two and what sorts of businesses do you advocate you know, the community model, but but maybe not product led growth? Or or, or maybe the better question is, what sorts of businesses lend themselves well to product led growth?

24:50
Yeah, no, look, I think make community slash open source, and product lead growth are kind of parts of the same continuum, honestly. I think community led growth is a necessary but not sufficient condition for success, right? community is a popularity metric, product lead growth is a profitability metric, right. And I think you need both. Now, you can use a community motion, or open source motion to become popular, right. But unless you do the hard work of building a good product that people find value in, you won’t be successful, right? It’s kind of what I tell my kids do. Like, yeah, you can be popular in school, but if you don’t pay attention to academics, don’t want to help you in the long run. So you know, I think, I think, early days, and by early days, I mean, like, you know, five, seven years ago, many companies fell in this fallacy of like, hey, if I’m popular, and I have millions of downloads, it’s guaranteed success, right. And you saw that in the early days of Hadoop, where you had cloud era where I was an investor and hortonworks, and, and map R, and then you had MongoDB, and couchbase. And many companies who had millions and millions of downloads, and downloads are great. It’s a vanity metric. But just because you’re popular, and people have good chatter around you, it’s not enough for you to drive value and is not enough for you to to drive revenue. So I think what companies learned the hard way is that communities are great. But until you could distill it down to a specific kind of time to value for your targeted user. It wasn’t enough, right? So you see a mom with a classic example where they realize that this gravy train of downloads then translate into support revenue was only going to take you so far, right? Once Amazon, Google, Azure said, Hey, we’re providing good enough support, you don’t need to go to Alaska, you don’t need to go to Cloudera, we’ll just kind of give you good enough support. In many ways, they raised the bar. And it became necessary for these community led open source companies to build a product lead motion, Mongo had to build Atlas, so that they could deliver Quick Time to value. It wasn’t enough to give you an open source database. That was cool. You had to go to a developer on your sass product, and help that developer write an application powered by MongoDB application that was operational within 90 minutes. If you’re then do it, people didn’t care how popular you were, they weren’t going to pay for it. Right? If they just wanted a popular software, they would just go to Amazon and say, Hey, you support it is good enough for me. Right. So I think that product led motion is a graduation from the community led motion, the best community led companies created great top of funnel great awareness. But the best companies then took it until the product lead motion that enabled the free users to try the free SASS product, get enough value, and then convert to the paid product. So I think it’s a necessary step in the evolution of this community led companies. And the best companies will look at open source companies like databricks MongoDB, elastic hashey Corp that started with open source community roots, has succeeded primarily based on their product led motion and their multi tenant cloud product, which delivers clear and distinct value. And frankly, it makes it harder for the cloud providers to catch up. Because if you have the community jobs and the brand, and you have a full functioning SASS product that delivers value, it’s very hard for a cloud provider to give you that, that mindshare and the product to compete with you. So I think it’s a competitive edge. And it’s a necessary step to have the product like motion to succeed.

28:24
Dharmesh, you’ve said that managing churn and focusing on customer success is more important now than ever. I don’t think many would dispute that. He, in what ways? Have you observed leading tech companies that are really applying innovative approaches to customer success? You know, whether it be expansion or retention?

28:43
Yeah, you know, it’s it’s surprising, I’ll tell you like, it’s so obvious the way you put it, that customer success is critical. And yet many companies I go to it’s like basic questions like, when do you start focusing on customer deployment, you know, you should get 80% of the customer purchase order deployed in the first 90 days after signing the deal, not when the customer is up for renewal, because by that time, it’s too late. Right? It’s kind of a reactive measure. It’s too late, right? And yet, you’ll be surprised how many companies I think the money’s in the bag, once a bill is signed, and they’re worried about customer success within 90 days of renewal. And by that time, it’s too late, right? I think companies that have the foresight, incentivize the customer success teams to be involved in the BOC cycle handoff smoothly to the customer success team, and ensure that you know, 80-90% of the deployment is up and running within 90 days, because that gives you a chance to upsell before the contract is due for renewal. Right? So that’s kind of like one on one in terms of the evolution of customer success. And I think the second and major change that we see happening now is this transition from subscription billing to consumption billing, right? Just like we went from perpetual software to subscription billing, we can now go in from a subscription billing to consumption billing where companies like snowflake and data, bricks and Mongo, and many others are saying, hey, pay us for what you use, you can subscribe to an annual deal, but you only pay what you use for. And in that case, customer success is basically in the expansion path to recognizing revenue, right? If you don’t deploy the software, customers are not going to pay for you for it, and you’re not gonna be able to recognize revenue. So you could have $100 million building all I care, but you won’t be able to recognize any of it, unless Customer Success is driving consumption, right. So for example, in a portfolio company, databricks, the entire company is focused on this dollar databricks usage per dollar, databricks units, customers committed to a contract. But unless they use those units, nobody’s getting rewarded, neither sales nor customer success, nor engineering, and the company doesn’t grow. So everybody is essentially a customer success person in a consumption model. And I think there’s increased awareness that customer success is offense, not defense. And I think that’s a one way path which benefits the end customer. And I think it enforces discipline on the software vendor, it puts the burden of success on the software vendor, not in the customer.

31:15
The feedback loop is much faster, right? When the usage goes down, that revenue goes down. And you’re, you’re raising the flag and saying, what are we doing?

31:25
Absolutely. And if we look at it, you know, Snowflake’s, done an exceptional job, I mean, the, the reason that they’re net dollar expansion on $1, retentions, north of 160%, is their focus and consumption. And that allows them an opportunity to upsell the customer, way before the contract is up for renewal. Because the more units you use up, not only do you want to renew with more units faster, but the next time around, you’re going to go up and commit to a much larger amount. So you don’t end up, you know, you don’t end up using all the units, you know, quickly. Right. So I think it’s a win win for the software vendor and the the end customer. Again, I think in many ways, the cloud providers have upped the game, Amazon, Microsoft, Google, they’re being paid as the services are being consumed, their sales reps, are making their orders or services are being consumed. And they have created good habits for the end customer. And they’re demanding the same from their software vendors. They don’t want to be a software vendor running on Amazon, a guaranteed monthly payment might not being the underlying provider a consumption based pricing, you’ll see that alignment. and that in turn puts customer success in the driver’s seat and tied to the hip itself. In terms of being on the offense side.

32:32
Should startups price that way from very early on. I mean, there’s always this question, you know, seats, instances, usage, remainder service…

32:43
…and defense, right. I mean, we talk a lot about like pricing markets, we have this cloud entrepreneur playbook that we talk about every year in terms of different aspects of building your business, in a bottoms up fashion, and we have this whole section on pricing market fit where I don’t think you can universally say everybody should go to a consumption model. It really depends on what the perceived value is, right. So for example, in that whole example of you know, engineering productivity, where if you’re using Atlassian, JIRA, or get lab or JFrog to make developers productive, the perceived value is the number of seats, because the more people use it, the more productivity benefit you get. And in that case, you’re perfectly comfortable paying, you know, whatever, hundreds of dollars per developer to save them 5-10 percent of their their time. And in that case, you can’t really base it on a consumption model. It’s a seed based license, right. But on the other hand, when you’re focused on data analytics, for instance, the more data you mined, in Snowflake or Databricks, the more value your business is deriving. And in that case, it’s tied to the volume of data, there, a consumption model makes a lot more sense. Because you want to incentivize the customer to use more data, you get more value, by using more data from the vendor. And in return, you pay them more for the value you receive. So I think pricing market fit is all about aligning the perceived value with your pricing unit. So you know, this consumption model works well, for volume oriented, compute oriented, software vendors, less so for seed based pricing models. And then you look at like security companies like security companies are often focused on an insurance and being proactive. I think they’re, you know, often it ends up being a seed based license, for example, aka you might pay based on the number of employees that are signing in, you can make that a consumption based model. So it depends on the type of company.

34:37
Do you believe should the customer success leader report up through the CRL?

34:44
I think so. I think so. I mean, a CRM is ultimately, let me step back. When I think about a simple structure of a public company, the CEO supported by the CPO, the product officer who runs all engineering technology, Product Management and CRO runs everything outbound In the CFO, of course, and during that model, their CRM is responsible for marketing, sales, customer success, anything and everything that impacts our customer lands on your product and expands over time is in the CRM function. So in that case, I think Customer Success becomes a part of the CRM organization. But I pay attention to how the individual customer success reps are incentivized, right? You don’t want to make them transactional, right? sales guys are the ones with the bad cops asking for the POS to be sign customers successful are the ones who are looking for the customer success. That’s the title in mind. And they’re often focused on getting the customers up and running. And by designing the compensation mechanism between CES and sales, to not be transactional, you have a transaction element, along with the deployment element. And that’s kind of how we can get the best outcome for for the this CRO’s organization

35:54
Would love to get transition a bit and get some of your thoughts on open source. You know, it’s heavily related to cloud, of course, but um, you know, what are your thoughts on open source and licensing models? You know, in recent years, there’s, there have been some impassioned and different points on this, and some are worried about integrity, open source, you know, others cast cast blame on the large tech companies like Amazon that are able to kind of monetize r&d that they didn’t create themselves, you know, where do you land on this? And sort of, what’s your position on investing in open source?

36:31
Two questions, right. One is investment thesis in open source, which I’ll cover in a second. And the other one is open source licensing models, which is kind of a touchy topic and important topic. But I think, I think open source company should just stop whining and focus on delivering more value to customers, like, you know, five, seven years ago could build a massive open source community. It was either because you know, your parent company allowed you to create this project and you know, create popularity on their dime, you look at Purdue came out a Yahoo and Google was behind it. And gasket came out on LinkedIn, and bark came out of amp lab. And they did a remarkable job building the popularity. But as I mentioned, popularity doesn’t necessarily translate to revenues and profitability, unless you build a cloud product, right. So in many ways, Amazon, Google, creating good enough support for open source really raised the bar for open source companies to get their next act. If Amazon and Google had not really pushed it, open source companies would be pretty happy setting support contracts, and you know, making 10s of millions in revenue. But the fact that Amazon Google pushed it, ultimately raise the bar and force the best companies to go create the cloud product to create a product lead motion to create value for the customers. And that’s what has resulted in the good open source companies that were proactive, like Mongo, elastic data, bricks, and you know, confluent, hashicorp, many others that saw this coming. And they increased their focus on their cloud product and implemented product lead motions. And there are others who kind of combining and today in order to change the licensing model, that was kind of a stopgap thing that gave you like a year or two, to block others and to get your act together. But if you didn’t get your act together, and then build a cloud product, you know, customers aren’t going to be just for support, right. So I think at the end, this was friction that was caused by cloud providers and software vendors. But at the end, it benefited the end customer. And it really separated the stronger open source cloud companies from the rest, who then pay enough attention to the cloud product. So you know, I think it’s a necessary element of evolution. And it creates a healthy dynamic, and nobody’s complaining here, right? I think snowflake has done exceptionally well as Amazon and on paper that compete with each other. But they’re both doing just fine, because the both are delivering specific value to customers. And the same is true for databricks. Same is true for, for Atlassian, and many other companies. So I think it’s, it’s a good healthy exercise, which raises the bar and the end customers benefit, right? As it relates to our investing strategy. When I think of open source as a customer acquisition channel, it’s not a business model, right. So if a company comes to us and says, our business model is open source, that conversation ends in 10 minutes, maybe five minutes. Notice on zoom, it’s much easier. But you know, I think companies that leverage open source to acquire, you know, solid top of funnel, and then use that to create a cloud product and take their open source users and downloads to a slack community where they engage and talk about the product, to then create a call to action to try the free SAS product or the free cloud product. And then use that to use some paid features and convert that user journey we absolutely love and like we cracked those metrics down, you know, having been part of many of the open source companies so we love open source companies that use open source or the customer acquisition channel. But then complement that with a cloud product and product led motions to acquire customers and build a business in a bottoms up fashion.

40:09
Love it. Dharmesh, you’ve been investing in cloud since I believe 2008. You know, what stands out to you, as you look across the big winners, any common threads or key success factors.

40:20
Yeah, no, I mean, the risk of not wanting to repetative. Again, I would say it’s just relentless focus on the the end user experience, right. Technology for too long, was focused on features, functionality, sophistication of the platform. And at the end, none of that mattered if it didn’t provide clear, concise value to the end user in a quick timeframe. And so, if I look at the last, whatever, 10-11 years of investing, I wrote my first check in 2011, the companies that have succeeded over time have continued to evolve, and continue to focus more and more on a well defined persona, very quick time to value and then matching the growth stage with the right go to market model. If you’re landing customers with 10-15k deals you need inside sales motions, which are focused on velocity and reducing any blocks. Once you land a bunch of customers that are fortune 500, global 2000, that’s when you layer and enterprise sales team to go expand and drive a higher net dollar retention. So I think a focus on the end user and delivering value, rather than just being carried away by technology and features has been the recipe for many companies to be successful. And it’s important to realize that technology and product are not the same thing. A lot of money in the.com days was built on technology, you know, a promising technology to enable faster internet, you know, optical multiplexing, what have you, the last decade has been focused on product, I don’t care how great a technology platform is, and how scalable it is. If you have a well defined product, that has been a recipe to succeed. And it sounds really simple when you say it. But it’s really surprising because so many people have succeeded in Silicon Valley, especially back in the late 90s and early 2000s. Based on building great technology, that is very hard for them to unlearn that, and relearn how to build a great product. And we often see founders who had an amazing track record in the.com era. And they had a very hard time adjusting to this new era where they had to build product, not technology. And if you look at the CEOs and founders of companies that have created more than 20 billion in market cap in the last decade, you’ll find many of them to most of them are first time founders that were not encumbered by legacy with a relentless focus on product. That’s been kind of my learning the last decade.

42:47
Love it. Love it. We talked with the team here about how there’s, there’s bright shiny objects everywhere, whether they’re technologies or even people, and it’s our job not to get, you know, not to head turn.

43:01
Indeed.

43:02
Dharmesh, what do you know, you need to get better at?

43:05
What do I know, man, I just got a lecture from my wife yesterday. 20 things. So I’m kind of go down that path. There’s a lot of things I need to get better at. But professionally, I think there’s a need to like, innovate and reevaluate a strategy a lot more often than ever before, right. So things are changing. So fast. Now, like in the last decade, we’ve seen so many changes from the cloud and data security and a Europe and Israel and India and open source cloud, and you name it, there’s so much going on, right. And if you say hey, at the end, our job is to find those point 1% of founders and point 1%, a company are going to generate 90% of the equity value or the next decade, you can’t get complacent. And what I need to get better at is questioning our strategy or assumption every six months every year and being prepared to change, right, because what was true last year, and even before COVID, we were worried about company shutting down in April. And then in September, more companies went public than and you know, like all of 2019 or something to that effect. And now companies are valued at like 100 billion at a faster pace than ever before, you seem like three or four companies go out at 100 billion. So, so much is changing. I kind of feel like many of our peers gonna get complacent. And many people are still like innovating in their offices waiting for their friends to call them to start a company and fund them. Whereas I think it Battery,we generally a lot more like eager to constantly learn. And I’m going to be ahead of the next thing. But I think the frequency has to change. You know, so five years ago, we saw Europe was a great source of innovation for b2b companies, and then hopped over to the US and we kind of went early in and found on data Iku and data org and mendix and believer and every company as well. We started 15 years ago before it was popular. New geographies open source we’re early on, you know, we kind of saw this consumption model and cloud business models And relentlessly focused on on that motion controllers cloud native, I think there’s just a lot changing. And what I need to get better at is constantly updating our strategy and, and challenging our assumptions. To make sure we’re ahead of the next big thing, because the bar is high, there’s a lot of really smart people in the venture ecosystem that have access to, you know, a great deal of dollars of capital. And smarts is not necessarily like a differentiator, but your investing strategy, the product market spaces you invest in and having a unique point of view, in the areas you invest in. It’s kind of the, the the driver of generating alpha in the space. And so I think I need to get better at at updating us.

45:45
And finally, here, Dharmesh, what’s the best way for listeners to connect with you and follow along with Battery?

45:50
Easy, their dharmesh@battery.com or LinkedIn. Those are the easiest ways to connect with me. And I’m always open to great ideas. And thanks again, Nick, for having me here. It is such a pleasure to be on the show. I’m a big fan of what you guys put together. And it’s an honor to be here with you.

46:06
Very much appreciate the time. Learn. Some have learned quite a bit every time we chat. And I’m sure the audience is going to really enjoy this one. So thank you, Dharmesh.

46:14
Nick, thanks so much. Appreciate it.

Transcribed by https://otter.ai