Datadog update (2/14/25)
Great Q4, bad reaction, what gives?
DISCLAIMER - This is not investment advice and you should do your own due diligence. I make no representation, warranty or undertaking, express or implied, as to the accuracy, reliability, completeness, or reasonableness of the information contained in this report. Any assumptions, opinions and estimates expressed in this report constitute my judgment as of the date thereof and is subject to change without notice. Any projections contained in the report are based on a number of assumptions, and there is no guarantee that projected outcomes will be achieved. I may, and likely do, have positions in the stocks I write about and may adjust or exit my positions without notice. I am not acting as your financial advisor or in any fiduciary capacity.
Datadog fell 8% after announcing earnings yesterday. This was despite a great Q4 print including a revenue of $738M that topped my own $736M estimate, because of a weak 2025 guidance below Street consensus on revenue and operating margins, spooking investors. It’s worth a quick sanity check with the data we have to see if what I wrote roughly a month ago holds, or if the story is coming apart.
Q4 Results Summary
Q4 revenue of $737.7M (25% y/y growth) and $3B+ ending ARR
Dollar-based net retention rate (DBNRR) ticked up a few points to high 110%s
AI-native ARR remained 6% of ARR (I estimate as roughly $180M), and contributed 5% of y/y growth (~$30M1)
Sales efficiency remained over 1x on an net new ARR / S&M spend basis
Non-GAAP gross margins ticked up to 82%, while FCF for the quarter hit 33%
Added 800 customers to get to 30k total, and added 66 $1M+ ARR customers in 2024
Guidance that made investors sad
Management guided 2025 revenue to $3.175-$3.195B, or $3.185B at the midpoint. This fell short against consensus estimates of $3.24B by about 2%, and my own estimate by 3%. It’s worth talking a bit about how the guidance game is played so we know what this means.
Most management teams like to start off with a conservative guidance that they can “beat and raise” across the year. By starting lower, management gives investors an implicit guarantee of what performance they’re going to deliver. Because everyone knows this is the rough floor, they will penalize the stock heavily for any miss or guide downward, but also expect management to continually come in above it. That’s the game, with the biggest unknown being how much the initial hedge is. Here’s how Datadog has done in the last 5 years:
The last two years are the ones most relevant for us. In 2023, it was very clear the macro environment was getting worse as interest rates went up to combat inflation. Despite the abrupt slow down in revenue growth, Datadog still beat its first revenue guide by 2.3%. In 2024, you could argue things got slightly better, and they came in ahead by 4.6%.
I’ll talk about the reasons they’d hedge later, but let’s assume for now they have a reasonably high degree of confidence of delivering a 2.3 - 4.6% beat for 2025. That would imply a full year revenue range of $3.26B to $3.33B, representing 21-24% growth instead of the paltry 18-19% in their guidance.
One other reference point is looking at Q4 2023, when Datadog posted 26% growth, then guided full year 2024 to 21% growth (which we’ve now seen ended up being 26%). At that time, consensus was expecting 23%, or 2% above guidance, which honestly probably should have been my clue of management’s willingness to guide below. My gut read is that management knew there were factors that could meaningfully accelerate growth, but even absent some of the AI native contribution, likely could have delivered 24% growth.
Finally, another way of looking at revenue is looking at billings coverage. Billings represents what they’ve charged customers, in both recognized and future revenue2. Since Datadog has a usage pricing model, billings will always trail actual ARR, but is one of the metrics we can actually see. In 2023 and 2024, Datadog had $1.85B and $2.35B in billings entering the year, or roughly 87-88% coverage on the final revenue they delivered. If we assume a proportional rate of growth and similar coverage to 2025, then ending billings of $2.88B also suggest around $3.3B in revenue.
I talk about sales efficiency a lot in my write ups, because efficiency actually tells you a lot about the health of a company, especially when you benchmark it against other SaaS companies. There’s also a lot of noise from quarter to quarter, but it gives the sharpest signal of when a company’s growth is seriously in trouble, regardless of what the top line results are. Q4 didn’t show any signs of major degradation, making me skeptical guidance needed to be this low.
I glossed over operating margins last time, but I’ll go over it quickly here since guidance for 21% margins also fell below Street consensus of 25%, and why I think it’s a non-issue. Once more, let’s look at what Datadog did a year ago - they guided to 21% margins, slightly below consensus at 22%, and ended up delivering 25% this year. There are lots of moving pieces with operating margin as well, but there are two primary reasons a company beats: 1) revenue out-performance over initial guidance, with the gross margin from the incremental revenue falling to the bottom line, and 2) spending less in operating expenses than you forecast. Here’s a table to make my math a bit more concrete.
For 2025, I’m putting a more conservative $3.28B as my revenue estimate and setting the opex cushion to $31M, same as it was in 2024. If you buy into management approaching the year similarly, you can see actual margins are only slightly going down.3 You could argue that perhaps the Street was also implicitly expecting Datadog to beat 25% and get closer to 26-27%, but as I said in my write-up, a few points here and there don’t bother me so long as the business stays intact. Datadog is investing ramp up sales capacity and win new logos, and building out new products to try to deepen customer relationships.
So why hedge this much?
I can’t get inside management’s head, but my best guess is they see more volatility in the core growth contributor for 2025, AI natives, and want to hedge as much as possible against further churn or attrition this year. Although other companies have an optimistic outlook on 2025, I can appreciate management not wanting to get caught off guard in trade wars or other geopolitical challenges make the year look more like 2023 instead of 2024.
Specifically on AI natives, commentary on the call pointed out that renewals saw optimization and discounts. Optimization is a nice euphemism for churn (you optimize by prioritizing your most important infrastructure/apps), although in Q&A the CEO Olivier mentioned higher commitments which I take to mean the overall contract still got larger, even if some of the observability products used by customers like OpenAI were discounted in the short term. Olivier also conceded it’s possible they lose someone like OpenAI entirely during Q&A, but it does feel along the lines of optimization that the most likely outcome is the footprint just gets heavily downsized.
At the end of the day, Datadog trades at a high multiple and is held to a high standard. If you buy my argument that margins aren’t materially off, then better growth in 2025 is what drives upside here. If revenue comes in around $3.3B, it wouldn’t be a bad year, but I’d understand investors who would ascribe a lower premium to that growth trajectory, leading to more modest price appreciation across the year. After all, that sort of result would mean AI inferencing has not yet translated into meaningful revenue for Datadog, and I’ve talked about how AI has been one of the clear drivers for upside with Datadog with broader cloud growth tapering off.
Personally, I think it’s too early to write off that potential and I still see a viable path to the $3.4B I called a month ago, although volatility with some of its AI native customers may make that figure more of a high watermark vs. baseline target. Deepseek’s impact has only really started to take effect in the last few weeks, and I can appreciate management taking a cautious stance and not building that into guidance. Yet if you’re excited for what new models could do for other companies like Cloudflare on the compute layer, it feels foolish not to think companies would tap Datadog’s products for the observability layer. I see Datadog playing aggressively to win more share in a crowded observability market. Even with moderate success outside of observability ($200-225M of ARR across the other products), pole position in observability remains such a huge prize that I remain bullish on their prospects.
I had estimated ~$33M in my write-up, and I’d be feeling really good about nailing Q4 revenue almost perfectly…if not for the stock dropping so much. The other point worth noting is with usage companies, ARR is less linear in the revenue recognition as it is with subscription, and unfortunately you can’t guarantee ARR / 4 is revenue in the quarter.
Here I’m looking at current billings, or revenue + change in current deferred revenue.
If your head is spinning trying to follow the math, don’t worry too much about it.