Latest / The CTO Podcast with Fexingo: Technical Leadership, Architecture, and Engineering Org / How Datadog Turned Observability Into Revenue
Transcript
- Lucas: I want to talk about Datadog because they have done something incredibly subtle but financially massive over the last eighteen months. They stopped selling just dashboards and started selling a unified data fabric for engineering teams. Luna: You mean the move away from pure monitoring into what they call full-stack observability? That sounds like marketing speak until you look at the retention rates. Lucas: Exactly. And the retention rates are where the story gets interesting. Most SaaS companies try to upsell features. Datadog tried to upsell context. They realized that if your engineers are looking at logs in one place, metrics in another, and traces in a third, they are losing time. So they built the pipes between those systems and charged for the convenience. Luna: That is a classic wedge product strategy. You give them the logs, they get addicted, then you sell them the metrics and traces. But how do you prevent feature bloat from slowing down development? Lucas: They avoided bloat by focusing on integration depth rather than feature breadth. Instead of building every possible visualization themselves, they made it easier for other tools to plug into their API. This turned their platform into a hub. By the end of the third quarter in twenty twenty six, their active integrations had crossed a significant threshold that competitors were still chasing. Luna: So the moat isn't the code anymore. It's the network effect of all these connected tools. That changes the competitive landscape completely. Lucas: It does. And it shifts the sales conversation entirely. Previously, a CTO would buy Datadog because their team was drowning in alerts. Now, they buy it because it reduces the cognitive load on senior engineers. That is a much harder metric to argue against in a board meeting. Luna: Cognitive load as a line item. I like that. But does that actually drive faster expansion revenue compared to traditional seat-based licensing? Lucas: It drives net dollar retention above one hundred and ten percent, which is elite territory. The mechanism is usage-based pricing on top of the base contract. When an engineering team scales up or adds new microservices, the bill goes up automatically without a new sales cycle. That is the holy grail for predictable revenue. Luna: Usage-based pricing always feels risky until it isn't. Did they face pushback from finance teams who hate unpredictable costs? Lucas: Initially, yes. But they introduced budget caps and forecasting tools that gave finance teams control. Once they solved the predictability problem, the engineering teams loved the flexibility. It became a self-serve growth engine inside large enterprises. Luna: Self-serve growth is the key phrase here. How did they onboard thousands of developers without drowning their support team? Lucas: They leaned heavily into documentation and community-driven troubleshooting. Their public status page and extensive library of pre-built integrations allow users to get value within minutes of signing up. It creates a viral loop where one developer invites five others to collaborate on a dashboard. Luna: That reminds me of the Slack early days. Bottom-up adoption followed by top-down procurement. Is that the model they are replicating? Lucas: Very similar. But with a longer sales cycle because enterprise security reviews are mandatory for observability platforms. You cannot just install a tool that sees all your code execution. That adds friction, but it also raises switching costs once you are in. Luna: High switching costs are good for valuation but bad for churn sensitivity. If prices go up, do they lose customers? Lucas: They raise prices incrementally and tie increases to value metrics like volume of data ingested. Since data ingestion correlates directly with system complexity, customers usually accept it as a sign of maturity. It feels fair rather than predatory. Luna: Fairness is a powerful driver in B2B sales. Do you think smaller competitors can survive this kind of ecosystem lock-in? Lucas: Only if they niche down hard. Generalist platforms will keep consolidating. Specialized tools for specific languages or industries might survive on the edges. But the core observability layer is becoming a utility, like electricity. Luna: Utilities don't need marketing departments. They just need reliable grids. That is a scary thought for any startup trying to build a new monitoring stack right now. Lucas: It is. Which is why I think the next wave of innovation won't be in monitoring itself. It will be in anomaly detection and automated remediation. Moving from passive observation to active intervention. Luna: Active intervention sounds dangerous. Who takes responsibility when an automated script breaks production? Lucas: Liability frameworks are still being written. But the market demand is clear. Engineers want to sleep at night without getting paged for routine fluctuations. Automation promises that peace of mind. Luna: Peace of mind sells. Especially to executives who haven't slept well since the pandemic. Does Datadog have a direct answer for this yet? Lucas: They have beta features for auto-remediation, but they are cautious. They position it as assistance, not replacement. Human oversight remains central to their philosophy. That caution probably protects them from catastrophic PR failures. Luna: Prudence wins out in enterprise tech. One major outage caused by AI could tank a stock price overnight. Risk management is part of the product design. Lucas: Absolutely. And that brings us back to the broader theme. Technical leadership is no longer just about building fast. It is about building safe, observable, and integrated systems that scale without breaking trust. Luna: Trust is the new currency. If your systems are opaque, you lose that currency quickly. Visibility buys you credibility. Lucas: Visibility buys you credibility, and credibility buys you budget. It is a virtuous cycle that rewards companies like Datadog who prioritize clarity over chaos. Luna: I wonder if other verticals will follow this pattern. Will CRM or HR platforms become equally integrated and usage-based soon? Lucas: Probably. Every complex software category is moving toward platformization. The winners will be those who make the underlying data accessible while keeping the interface simple. It is a difficult balance to strike. Luna: Simple interface, complex backend. The classic paradox of modern software engineering. It seems like the only way to win is to embrace that complexity rather than hide it. Lucas: Precisely. Hide it, and you build a fragile house of cards. Embrace it, and you build a foundation that can support whatever comes next. Luna: That is a good principle to take away. Whether you are building a platform or just managing your own team's workflow. Lucas: Agreed. Complexity is inevitable. Transparency is optional. Choose transparency. Luna: Thanks for breaking that down, Lucas. It makes the whole platform strategy feel less like magic and more like disciplined execution. Lucas: Anytime. Before we wrap, I want to mention something real quick. A small group of listeners chip in monthly through buy me a coffee dot com slash fexingo — that's F. E. X. I. N. G. O. — and that literally funds making these deep dives possible. Luna: Yeah, it keeps the ads off and the research thorough. I appreciate everyone who supports the show directly. Lucas: It really does. Looking ahead, I am curious if you think the definition of a CTO will change as these platforms become more autonomous. Luna: I think it shifts from builder to architect of trust. Less coding, more governance. But that is a story for another day. Lucas: Sounds like a plan. Until then, keep observing carefully.