Latest / Startup Stories with Fexingo: Conversations About Founders, Funding, and Building Companies from Zero / How Datadog Built a $50 Billion Company From One Log
Transcript
- Lucas: Look at the chart for Datadog right now. It’s not just that they hit fifty billion dollars in market cap; it’s how they got there without hiring a thousand sales reps. Most enterprise software companies, you think about them as consultancies disguised as tech firms. They send guys in blazers to talk to CTOs for six months. Luna: Right, because traditionally, selling to IT departments is a nightmare of bureaucracy and long procurement cycles. Lucas: Exactly. But Datadog flipped it. They started with one thing: logs. Just text strings coming from servers. And they made it so easy that developers installed it on a Friday afternoon, forgot about it, and then the company was paying for it by Monday. That is a radical shift in who actually buys the software. Luna: It feels like they understood that the person using the tool isn't always the person writing the check, but the user has veto power over everything. Lucas: Precisely. Olivier Pomel, the co-founder, kept saying that if you make life easier for the engineer, the finance department eventually notices the bill. By October twenty twenty-six, that model is proven across half the Fortune five hundred. But I want to look at the specific mechanics of how they expanded from logs into what they call observability. Luna: And whether that expansion diluted their core value proposition or just deepened the moat. Lucas: So here is the thing about early cloud infrastructure. Before Datadog, if your server crashed, you were staring at a terminal window filled with gibberish. You had no idea where the error came from. It was like trying to find a needle in a haystack while blindfolded. Datadog aggregated all those logs into a single searchable interface. Luna: It sounds simple now, but back then, setting up log aggregation required its own team of specialists. It wasn't plug and play. Lucas: That friction was the entire business case. They removed the setup cost. Once you have the logs, you naturally start wanting to see the metrics—CPU usage, memory spikes—and then traces, which show you how a request moves through microservices. They didn't build three products. They built one platform that just saw more data. Luna: I remember reading about their pricing model being based on ingestion volume. Does that create a ceiling for their growth? Lucas: That is a common critique. If you pay per gigabyte of data, aren't you capping your revenue by encouraging customers to delete old data? The counter-argument is that retention and compliance require keeping that data longer. Companies can't afford to lose visibility just to save ten percent on the monthly bill. Luna: Yeah, especially when an outage costs hundreds of thousands of dollars an hour. Cheaper to pay for the data than to guess during a crisis. Lucas: Let's talk about the community aspect for a second. This is where the 'product-led' part gets interesting. They built a massive public blog and educational content specifically for sysadmins and developers. It wasn't marketing fluff; it was technical tutorials on how to debug Kubernetes clusters. Luna: They essentially trained the workforce to use their tool before the workforce even knew they needed to buy it. Lucas: Spot on. Students learn on Datadog in university labs. Junior engineers join a company and say, 'We should use Datadog because I know how to query it.' It creates a bottom-up pressure that bypasses traditional IT procurement entirely. Luna: It’s fascinating how much leverage a single tool can give a junior employee in a large organization. Lucas: Before we go deeper into the numbers behind their international expansion, I want to pause for a quick moment. These conversations about building companies from zero are only possible because of listener support. If these insights have moved your work forward, even in a small way, consider helping us keep this network ad-free and independent. You can support the show directly at buy me a coffee dot com slash fexingo — that's F. E. X. I. N. G. O. Luna: It really does help us dig into these case studies without needing corporate sponsors dictating the angles. Lucas: Agreed. So, let's get back to the international question. In twenty twenty-three, they opened offices in London and Paris, but the real test was Asia-Pacific. How do you sell a us centric cloud tool to a region with different regulatory concerns about data sovereignty? Luna: Data residency is huge in Europe and increasingly in Asia. You can't just host everything in Virginia anymore. Lucas: Datadog responded by building local regions. They allowed European clients to store their telemetry data within EU borders. It’s a subtle operational change, but it signals respect for local laws. Without that, you lose the enterprise deals in regulated industries like banking. Luna: It’s a classic example of adapting the infrastructure to fit the customer, rather than forcing the customer to adapt to your infrastructure. Lucas: Exactly. And it ties back to that developer trust. If your tool breaks compliance, your engineers won't use it, no matter how pretty the dashboard is. Now, looking at the current quarter, we see their gross margins holding steady above seventy percent. That is incredibly high for a data-heavy business. Luna: Which suggests they’ve optimized their storage costs significantly, or found a way to add value that outweighs the storage expense. Lucas: Both. They introduced ai driven anomaly detection. Instead of just showing you a graph, the system tells you, 'Hey, this spike is weird.' That requires heavy compute, but it increases stickiness. You don't uninstall a tool that predicts outages for you. Luna: So the moat isn't just the data; it's the intelligence layer sitting on top of it. Lucas: Right. And that leads to a broader point about the market today. There are dozens of competitors trying to slice off pieces of that pie. New Relic, Dynatrace, Splunk. Why hasn't Datadog been eaten alive yet? Luna: Switching costs must be astronomical once you're fully integrated into your CI/CD pipelines. Lucas: Yes. And their integration ecosystem is massive. They have pre-built plugins for almost every major framework. Migrating away means rebuilding all those connections. It’s a utility. Nobody replaces the electricity grid just because a new bulb looks cooler. Luna: Utility status is a powerful position to hold, assuming the reliability holds up. Lucas: Reliability is non-negotiable. If the monitoring tool goes down, you’re flying blind. Datadog maintains a ninety-nine-point-nine-nine percent uptime SLA. That level of engineering rigor reinforces the brand promise. Luna: It seems like they bet heavily on engineering excellence rather than marketing spend. Lucas: Their marketing spend is actually quite efficient compared to peers. Because the product spreads virally among engineers, their customer acquisition cost is lower. They reinvest that savings into R&D, which improves the product, which drives more word of mouth. It’s a virtuous cycle. Luna: A flywheel effect driven by developer satisfaction rather than executive persuasion. Lucas: Now, let's look at the risks. As we move into late twenty twenty-six, interest rates are stabilizing, but growth expectations for SaaS companies remain high. Can Datadog continue to grow at twenty-five percent year-over-year without hitting a wall? Luna: Or will they face saturation in the mid-market segment where most startups already use them? Lucas: That’s the big question. The mid-market is saturated. The next frontier is the legacy enterprise. Banks, insurance companies, manufacturing plants. These organizations move slowly. Convincing them to adopt modern observability is harder than convincing a Series B fintech. Luna: Do they have the patience and the sales muscle to tackle those slower-moving giants? Lucas: They’ve been hiring account executives, but it’s a cultural shift. They’re moving from pure product-led to hybrid. Some purists worry this dilutes the brand. Others see it as necessary for survival at scale. Luna: It’s a balancing act. You need sales to close the big deals, but you can’t lose the agility that got you here. Lucas: Another angle is security. Observability tools see everything in your code execution. They become prime targets for attacks. Datadog had to invest heavily in securing their own platform. If they suffered a breach, the reputational damage would be catastrophic. Luna: Trust is their currency. Any leak in that vault undermines the whole business model. Lucas: Absolutely. So, looking ahead, the story isn't just about monitoring anymore. It's about AIOps—Artificial Intelligence for IT Operations. Predictive maintenance, automated remediation. Datadog is positioning itself as the brain of the infrastructure. Luna: From a flashlight to the central nervous system of the cloud. Lucas: That’s a fair way to put it. The valuation reflects that ambition. Fifty billion dollars is a lot of money for a company that started by indexing text files. But the margin for error is thinner now. They can't just coast on momentum. Luna: They have to keep innovating faster than the complexity they’re managing. Lucas: Which brings us back to the founders. Olivier Pomel still writes code. He’s involved in the product roadmap. That hands-on leadership style is rare at this size. It keeps the company grounded in reality. Luna: It prevents the disconnect between the boardroom and the bug tracker. Lucas: Right. In a world of detached CEOs, that direct line to the technology is a competitive advantage. It ensures they solve real problems, not perceived ones. Luna: Do you think other startups can replicate this model, or is Datadog too far ahead in the network effects game? Lucas: The model is replicable, but the timing was perfect. They arrived exactly when microservices broke monolithic debugging. Being first mover gave them the data network effect. Newcomers have to convince users to switch platforms, which is a hard sell. Luna: First mover advantage in infrastructure is incredibly sticky. Lucas: It is. But innovation never stops. We might see a new wave of tools focusing on edge computing or serverless functions, challenging Datadog’s dominance in a different way. Luna: The landscape is always shifting under their feet, even if they’re currently on solid ground. Lucas: True. The question is whether they can adapt as quickly as the underlying technology changes. Their track record suggests they can, but the pace is accelerating. Luna: It’ll be interesting to see if their AI features become the standard or just another niche capability. Lucas: If they pull it off, they become indispensable. If they stumble, they risk becoming a commodity utility with thinning margins. Luna: High stakes for a company that started with such a simple premise. Lucas: Simple premise, complex execution. That’s usually the recipe for a unicorn. Luna: Or a cautionary tale if they lose focus along the way. Lucas: For now, the data points upward. But in tech, tomorrow is a long time away. Luna: We’ll have to watch their next earnings call closely to see if the narrative holds. Lucas: Agreed. Until then, we keep tracking the logs.