Shares of Datadog (DDOG -0.10%) fell on Tuesday, down as much as 6.3%, before recovering slightly to a 4.1% decline as of 12:51 p.m. ET.
Today's sell-off was caused by a negative analyst note, which posited Datadog's growth rate could take a big step down this year due to the loss of its largest customer, AI disruptor OpenAI, which is reportedly building its own observability software tools in-house.
OpenAI may look to cut out Datadog
Since the COVID-19 pandemic, Datadog has emerged as a disruptive winner in log management and observability software, which monitors the health and security of IT infrastructure. Datadog's cloud-first approach has found favor especially with high-growth tech companies, leading to a strong near-50% average growth rate over the past five years.
But today, sell-side analysts at Guggenheim downgraded Datadog's stock from "neutral" to "sell," while putting a $105 price target on the stock. That's significantly below the $146 price at which the stock trades today, even after Tuesday's negative performance.
The reason for the downgrade is Guggenheim's take that Datadog will lose its largest customer in OpenAI, the company behind ChatGPT, which is reportedly building its own in-house log management metrics software.
The analysts see this loss resulting in a hole of more than $150 million in revenue, which could result in a significant deceleration later this year. Guggenheim sees revenue growth stepping down from the current mid-20% growth rate to 17% in the fourth quarter of this year and 15% growth in 2026.

Image source: Getty Images.
The downgrade is worrying not only for Datadog
Recent news reports have highlighted that OpenAI has broad ambitions to take its market-leading large language AI models into numerous downstream applications, from enterprise software to even its own hardware devices. Since large language models are now fairly adept at writing code, could OpenAI also displace its own software suppliers?
To be fair, OpenAI's AI infrastructure likely processes orders of magnitude more data than the typical Datadog customer, which correlates to massive amounts of money currently going to Datadog. That could make it financially sound for OpenAI to build its own internal software, but not the rest of Datadog's customer base.
Yet while this won't be the case for every company, it does raise the potential of OpenAI and other AI-first start-ups becoming competitive threats to existing software leaders, as these new start-ups aim to go "downstream" into customer-facing applications.