To The Moon

When a great quarter becomes a bad result

I have always thought Datadog was one of those software companies where the business can do almost everything right and still annoy investors.

That peculiar talent was on full display after its August 6 results. Revenue jumped 36% year on year to $1.12 billion, beating expectations, while adjusted, non-GAAP EPS came in at $0.65 versus $0.58 expected. Management also raised full-year guidance for the third consecutive quarter.

When your biggest customer can build the competition itself

The reward? The shares fell 19% over the two sessions after earnings.

That is not the market declaring $Datadog(DDOG)$ a busted business. It is the market discovering, rather painfully, that when you pay a premium price, even an excellent quarter can arrive wearing the wrong shoes.

The flashpoint was weaker-than-expected third-quarter guidance. Management said its largest customer, a leading AI company with a nine-figure contract covering 17 products, would reduce usage from the third quarter. Wall Street widely believes that customer is OpenAI.

Suddenly, an old theoretical threat to software has acquired a very real invoice.

The warning was hiding in plain sight

The fascinating part is that this is not a brand-new bear thesis.

In July 2025, Guggenheim downgraded Datadog to sell, warning that OpenAI's development of internal monitoring tools could eventually remove around $150 million of Datadog revenue.

The market largely shrugged. Datadog then nearly doubled over the following year.

Now management has effectively confirmed that the underlying issue exists.

That does not mean Guggenheim's original worst-case scenario is coming true. It does mean investors should stop treating AI cannibalisation as an entertaining conference-call talking point and start treating it as a variable in the valuation model.

That distinction matters.

If AI-native companies become unusually aggressive about building infrastructure internally, Datadog could face a customer cohort that behaves differently from traditional enterprises. AI companies are not merely deploying more software; they are increasingly capable of building software themselves.

And that is where the wider 'SaaSpocalypse' debate gets interesting.

Datadog is trying to eat the problem

There is a rather clever counterargument buried inside Datadog's strategy.

The company is no longer positioning itself simply as the place where developers look when something breaks. Its product ambitions increasingly involve AI itself, including Bits AI and an agentic approach designed not merely to observe problems but help fix them.

That is strategically important.

If AI reduces the need for conventional monitoring, Datadog's answer is effectively to make the monitoring layer more intelligent and more deeply embedded in the customer's workflow.

In other words, if the robot is coming for your software, teach the robot to use your software.

That strategy is not guaranteed to work, but it is considerably more interesting than defending yesterday's product catalogue. And the underlying customer data gives management some ammunition.

The financial engine remains formidable

The spreadsheet shows TTM revenue of $3.97 billion, up 31.5%, compared with $3.43 billion in FY2025 and $2.68 billion in FY2024. Free cash flow reached $1.18 billion, up 27.7%, producing a formidable 29.8% FCF margin.

The revenue engine is humming; earnings are still playing catch-up

Operating cash flow was $1.23 billion.

Datadog also finished the latest period with $4.99 billion of cash and short-term investments against $1.28 billion of total debt, leaving net cash of $3.71 billion.

That is an enviable balance sheet.

Customer expansion is equally important. Datadog had 33,400 customers at the latest TTM period, while 85% used at least two products, 58% used four or more, 37% used six or more and 22% used eight or more.

That progression is one of the less obvious strengths of the investment case. Datadog is gradually becoming less of a single-product software vendor and more of a platform embedded across a customer's technology estate.

There is a catch, however.

Stock-based compensation reached $823 million over the trailing twelve months. That is a substantial figure relative to the $1.18 billion of reported FCF, while diluted shares outstanding have risen from 309 million in 2021 to 366 million.

I therefore like Datadog's cash generation, but I would not pretend dilution is merely a rounding error wearing a small hat.

Competitive analysis: the moat is real, but AI changes the map

Datadog competes with $Dynatrace Holdings LLC(DT)$, New Relic, Cisco's Splunk and a growing collection of cloud-native and open-source observability tools, while $Amazon.com(AMZN)$, $Microsoft(MSFT)$ and $Alphabet(GOOGL)$ can leverage their own cloud ecosystems.

Datadog's advantage is breadth.

The company's increasing multi-product penetration creates switching costs and gives it an opportunity to expand revenue inside existing accounts rather than constantly replacing churned customers with new logos.

That is a powerful model, but the competitive threat from AI is different from traditional software competition. A conventional enterprise may prefer buying a mature observability platform rather than assembling one internally. An AI-native company with enormous technical resources and unusual economics may not share that instinct.

That is why the OpenAI development matters beyond its direct revenue impact. It is a test case for whether AI-native customers will behave differently from the enterprise customers that helped build Datadog's growth model.

The wider software market is already behaving as though the second possibility deserves attention. HubSpot fell around 20%, while other SaaS names have also come under pressure as investors ask whether AI will compress software budgets rather than simply create new ones.

Datadog has consequently become something of an unwilling poster child for the argument.

The valuation is where the argument gets uncomfortable

At $233.93 on August 7, Datadog's market capitalisation was approximately $84 billion.

The spreadsheet puts the trailing P/E at 476.95x and forward P/E at 88.44x. The shares trade at 21.18 times sales and 71.06 times FCF.

Those numbers explain the sell-off better than any single guidance line.

When expectations slip, expensive stocks discover gravity quickly

A company trading at more than 70 times FCF cannot merely be excellent. It has to remain excellent for a very long time.

And Datadog's current profitability makes that especially demanding. TTM net income is only $177.58 million, with an operating margin of 0.4%, despite the impressive cash-flow profile.

The market is therefore valuing Datadog principally on what it can become rather than what it currently earns.

That can work spectacularly well. It can also become spectacularly unforgiving.

The real question is what happens next

I do not think the 19% decline invalidates the Datadog bull case.

Revenue is still growing at more than 30%. Customer product adoption continues to expand. Dollar-based net retention is approximately 120%. Remaining performance obligations are $3.47 billion, versus $2.27 billion two years earlier. The balance sheet is exceptionally strong.

But the OpenAI issue changes the conversation.

The most important question is no longer whether Datadog can continue growing. I think it can. It is whether AI-native customers will eventually grow their Datadog consumption as quickly as traditional customers have — or whether some of the industry's cleverest companies will increasingly decide to build around it.

The observer may soon discover it is being observed

That is a much harder question.

For me, Datadog remains one of the better businesses in software, but the stock is not priced like a business allowed to make many mistakes. At 88.44 times forward earnings and 71.06 times FCF, I would want considerably more valuation support before becoming aggressive.

The irony is that Datadog's latest quarter may ultimately prove healthy for the investment case. It has forced the market to acknowledge a risk that was previously theoretical.

Sometimes a 19% fall is not the end of the story. Sometimes it is the chapter where the story finally becomes interesting.

@TigerStars @Daily_Discussion @Tiger_comments @Tiger_SG @Tiger_Earnings @TigerClub @TigerWire