ServiceNow Stock Fell 15% After Beating Earnings. Here's Why (NOW)

ServiceNow beat earnings and fell 15% after hours

0:00 ServiceNow just beat Wall Street's revenue estimate and raised its full-year outlook for the second time this year. And the stock dropped about 15% after hours to around $127, which by the way, that's just after hours activity and it's very subject to change when the market opens tomorrow. So, if you've been sure all year that AI is going to eat software companies alive, this probably looks like your proof. But, if you've been waiting for a discount on one of the biggest software businesses out there, it might look more like a gift.

0:25 Full disclosure, I myself own about 100 shares of ServiceNow. So, tonight's drop hit my account, too, and I'll show you exactly what that looks like. The easiest way to picture what happened is your favorite restaurant posting its best month ever, packed every night, and then the owner mentions that the reservation book for next month is filling up a little slower than it used to. The food is great and the room is full, but people who own a piece of that restaurant mostly care about one thing, the reservation book.

0:50 That's basically ServiceNow tonight, a great month, but a slower reservation book. If you've never heard of ServiceNow, it's basically the software that large companies use to route work from one person to the next.

What ServiceNow actually sells (and why seats matter)

1:01 When your laptop breaks at work and you file a ticket or you ask HR question through some company portal, there's a good chance that request is running on ServiceNow. Companies pay for it like a subscription and for most of its history, they paid per seat, which basically just means per employee who logs in. And this year, ServiceNow has been pitching itself as what it calls the AI control tower, the place a company keeps track of every AI agent that's running and what those agents are allowed to do.

1:25 That's the exact problem ServiceNow says that it solves. But, here's the question the market has been chewing on all year.

The AI seat fear, the crash and the comeback

1:30 If AI agents can do the work, do companies still need as many people logging in? Fewer people logging in could mean fewer seats and fewer seats could mean less revenue. And that one fear hit almost every software stock at the end of January. ServiceNow was one of the hardest hit, falling from over 150 at the start of the year to around $90 by late June. The guy who sure everything goes to zero was telling anyone who would listen that software was dead.

1:53 But, then the stock came roaring back, up more than and from late July to the close on October 28th as investors started betting AI might actually help ServiceNow instead of killing it. The guy who bought near $90 looked like a genius and he made sure his wife heard about it at dinner. And naturally, the afternoon right before earnings is when Wojak finally decided to buy in. So, walking into this report, the stock wasn't priced like a company in trouble anymore.

2:21 It was priced like a company that had to be perfect. So, let's see whether it was perfect starting with the good stuff because there are a lot of it.

The good numbers from Q3

2:28 Subscription revenue, which is the money customers pay to keep using the software, came in just under 4 billion for the September quarter, up 21% from a year earlier. Total revenue was about 4.1 billion, a little ahead of the 4.09 billion analysts were expecting. Big customers kept spending big, too. ServiceNow signed 19 deals worth more than 5 million each in new annual contract revenue, and that count more than doubled from a year earlier. It now has 680 customers paying it more than $5 million a year, up more than 20% in a year.

2:56 And ServiceNow raised its target for new AI contract value this year to 1.7 billion, up from 1 billion it started the year with. Said another way, the AI products that were supposed to be a threat are turning into one of the fastest-growing things ServiceNow sells. On top of that, management raised its full-year subscription revenue outlook again and raised its free cash flow margin target by a full percentage point, partly by spending less on equipment.

3:19 If you only read that list, you'd expect the stock to be up. So, why did it fall about 15%?

Why it fell: the slower reservation book (cRPO)

3:25 Well, let's go back to that reservation book. ServiceNow reports something called current remaining performance obligations or CRPO, which basically just means contracted revenue the company expects to collect over the next 12 months. That's the reservation book, and for the December quarter, ServiceNow guided CRPO growth to 18.5% in constant currency, down from 21% this quarter. Constant currency basically just means the growth rate with exchange rate swings taken out since some customers pay in other currencies than USD.

3:51 A slowdown of a couple points doesn't sound like much, but for a stock priced to four perfection, a slow reservation book is exactly what the AI bears were waiting to see. On the call, the CFO said that CRPO is, quote, a very nuanced accounting metric and it does not necessarily capture all of the momentum in the business. And to be fair, she has a point. Because the full reservation book, including contracts that run past the next 12 months, grew 22% to about 29.7 billion.

4:18 But it's still a little like the restaurant owner telling his investors the reservation book is a nuanced metric. His investors did not find that nuanced.

EPS miss, gross margin, buybacks and acquisitions

4:25 The second problem was profit per share. Adjusted earnings came in at $1 against the $1.02 that were expecting. 2 cents is tiny, but the reason behind it matters more than the size. ServiceNow's subscription gross margin, which is how much of each subscription dollar is left after the cost of actually running the service, slipped to 80% from the low 80s a year ago. My read is that running AI isn't cheap, and when customers use a lot more AI, somebody has to pay for all those computer chips.

4:50 The company didn't spell it out in the release, so treat that part as my opinion. And the last thing that investors noticed is that ServiceNow did not buy back any of its own stock this quarter. Compared to almost 600 million a year earlier. Instead, it has spent about 9.2 billion dollars buying other companies so far this year, including the security company Armis. Buying companies can absolutely pay off, but acquisitions also add revenue on their own.

5:11 So it gets harder to tell how fast the original business is really growing. For example, ServiceNow guided December quarter subscription revenue to grow 22.5% to 23%, which sounds like a speed up from 21%, but in constant currency it's 21.5 and the company didn't break out how much comes from the companies it bought. And just like that, a big chunk of the comeback disappeared in one evening.

Good numbers, nervous reaction

5:37 And if this feels familiar, that's because it is. ServiceNow beat on revenue in July, too, and the stock still fell about 4% the next day. And after the April report, it fell almost 18%. So for most of this year, the pattern with the stock has been good numbers, nervous reaction. So is this proof that AI is eating ServiceNow?

Is AI eating ServiceNow? Bull case vs bear case

5:55 Well, here's where it gets interesting. On the call, CEO Bill McDermott said, quote, "The customer likes the predictability of the C-based pricing model for their budgeting purposes." And the CFO said customers are, quote, "not asking at all for outcome pricing." Which basically just means paying for what the AI gets done instead of for how many people log in. The bull case is that ServiceNow gets paid either way. It keeps charging for the humans and starts charging for the AI agents working alongside them.

6:19 McDermott's line was, quote, "ServiceNow controls what agents can do." And if that's true, every new agent a company turns on is one more thing it needs ServiceNow to keep an eye on. He also said, quote, "As the cost of intelligence falls, the tailwind for software is strengthening." But in the same call he said, quote, "Solutions that don't matter or point solutions, they will get eliminated." And a point solution basically just means software that does one narrow job.

6:42 So even the biggest bull on the call agrees some software gets replaced by AI. He just doesn't think it's his. The bear case is simpler. If a help desk that used to need 500 people only needs a hundred, that's a lot fewer seats, no matter how nice the control tower is. An honest answer is no one knows which one wins yet, including me. So what would I actually watch from here?

The 3 things I'm watching next

7:02 The first thing is that reservation book. If CRPO growth gets back above 20% over the next couple of quarters, the slowdown was a blip. The second thing is gross margin, because if it keeps slipping as AI usage grows, the AI business may be growing revenue faster than it grows profit. And the last thing is that 1.7 billion AI target, because hitting it would be real evidence that AI is adding to ServiceNow instead of replacing it. As for the price, around $127 after hours puts the stock below where it started the year, but still well above the low of about $81 it hit earlier this year.

7:34 And as I mentioned earlier, after hours prices can swing a lot by the next morning. So the real reaction is whatever the stock does once the market opens. If you're wanting fast growth with no drama, this probably is not your sign to buy. But if you think the AI fear is overdone, a great quarter with a slow reservation book is the kind of messy setup that sometimes looks cheap in hindsight. Or it doesn't, and that's the risk. As for me, I'm just holding my shares and I'll let you know if that changes.

Why I'm holding

7:58 The restaurant is still packed. The question is next month's reservations. Let me know in the comments if you think that AI helped ServiceNow or if it eats it. YouTube picked this next one for you, and unlike Wojack, it won't panic after hours.