Sometimes You’re The Datadog, Sometimes You’re The Hydrant

When the media covers companies created by technology, names like Apple, Uber and Slack come up. But what usually doesn't get covered is that behind a new generation of SaaS and social apps is a new generation of infrastructure tools and services that support them. You may not know what they do, but companies like Elastic, Fastly and Crowdstrike are just as important to the tech ecosystem as the ones you interact with every day.

We saw the same pattern in the growth of the industrial economy. 

Consider the chicken breast on your plate for dinner at night. In order to get the chicken from egg to oven, an entire supply chain had to be created. Feed production, feed storage, chicken production, transportation, cold storage, cooking, packaging, retailing and disposal — there are billion dollar companies at each step.

The information economy works in the same way. In service of producing and maintaining software applications, billion dollar companies have been built to:

• Accept online payments
• Secure data
• Help engineers debug code
• Enable users to login across applications
• Load images and videos faster
• Digitally collaborate
• Manage devices remotely

And many others. Though few of these touch the end consumer, they are big business. And Datadog is one of these software infrastructure companies. 

Specifically, Datadog monitors technology infrastructure and is used to prevent and react to technology outages or failures. If a user sees an error message or can’t use your software, that’s a huge problem. It needs to be solved quickly and efficiently: find the bug, make the change, and ship it across the system. Datadog helps companies do that. 

Datadog built software that tracks and monitors metrics, logs and traces for a system. Metrics are a numerical representation of data over a time period. An example might be the response time of a webpage. Logs are records of discrete events. Metrics can be made up of logs, but log data is more granular. It’s used for more detailed debugging. Traces are a series of events that are causally related and make up the end-to-end flow of a request through a system. It’s the path from when a user does something (like click play on a video or message their friend) and  how that action moves throughout the technology stack. 

Once Datadog collects all this data, they display it in a consolidated dashboard; a single pane of glass, if you will. It’s easy for everyone to see how the system is performing -- from developers to business analysts and everyone in the middle. Their product integrates across hundreds of different data sources, alerts users anytime a metric is outside the normal range, and makes it easy to identify the root cause when a problem surfaces. 

You might be thinking, hasn’t this always existed? Datadog was founded in 2010 and launched their first product in 2012. Yet, we’ve been building software for decades. Why have they done so well? 

There are two reasons: market and product. 

Datadog was in the right market at the right time. In the early 2010s, cloud computing proliferated. With it came two trends that fit Datadog’s strategy. Development and operations teams merged in corporations (“DevOps”) and thus a broader range of people needed to view infrastructure analytics. And the other trend was the increased use of microservices, containers, and serverless computing. This increased the complexity of the software stack and demanded a full “observability” solution. Datadog replaced software that was either built for the on-premise computing environment or was only a single-point solution in the cloud.

But it wasn’t just the market. Datadog saw that the cloud was the future and that the increased complexity demanded a single, consolidated solution. Horizontal, not vertical. Their product strategy, pricing strategy and business model all came from this thesis. They lead with an infrastructure monitoring product, then built out APM and logging to round out the “three pillars of observability.” The past two years they’ve launched an additional five products which have seen great early traction. 

Datadog has been one of the best performing SaaS companies in the world. They were built by developers, for developers, and sold in a way that developers loved (e.g., not sold to at all). Datadog has incredible bottoms-up sales motion which has resulted in dollar based net retention of at least 130% for ten straight quarters. 

And while Datadog spent the past several years leapfrogging ahead of their competitors, other monitoring software has become more comprehensive. Splunk and New Relic are just two examples of point solutions that have matched Datadog’s suite of products in the past couple years. Despite their incredible success to date, Datadog’s future as a stock is far from certain. 

Let’s dive into the details.

Cloud Computing

In 2006, Amazon launched Amazon Web Services. And with it came a better way of creating software. From a post I helped write about Rackspace:

It was the first large and accessible cloud computing platform. This seminal moment changed the enterprise landscape: before the cloud, businesses had in-house server hardware, software licenses, and a slew of IT professionals to manage it all. This on-premise infrastructure was inflexible and expensive. Cloud adoption made software architecture more flexible, cheaper to maintain, and easier to scale up and down.