Ok, but what about the spying?
And: Is tech buying slowing?
Welcome to Cautious Optimism, a newsletter on tech, business, and power.
Happy Wednesday! CO has never received so many texts and notes about a headline as we did yesterday since the internally infamous Sunshine and moonbeams, motherfuckers.
My last word on the present TechCrunch conversation is this: There are still incredible people there. Theresa and Maggie holding podcasts and video down. Karyne, the editing genius. Kirsten, the transport savant. Zack leading the cybersecurity team. Julie, Becca, and Dom on the venture beat. Tim’s tenacious climatetech coverage. The list goes on. Ozempic is not starvation. (Media layoffs just keep coming, however; it’s brutal.) — Alex
📈 Trending Up: Worries about domestic fiscal policy … public journalism … crowdsourced opinion? … AGI timelines … Cohere … Manus … behold, the captive mind …
📉 Trending Down: Bright lines between political power and personal business … public cybersecurity … lunacy … a slowdown in data center buildouts … ADP employment expectations for May …
Ok, but what about the spying?
Remember when HRtech and payroll giant Rippling accused rival Deel of hiring one of its workers to exfiltrate information about its business? Right, the spy thing that involved the man hiding in a bathroom trying to clear his phone, honeypots, and a lot of brutal tweets for Deel, which has raised hundreds of millions of dollars.
Time has passed. Now, Deel has fired back in a filing seen by the FT that Rippling was also spying on its own operations. We’ll learn more when we get our hands on the document in question, but Deel has paired that claim with an announcement that it “crossed $1 billion annual revenue run rate in the first quarter and has earmarked up to $500 million for acquisitions this year to bolster growth,” per Reuters.
Great. But what about the spying thing? Either Rippling is the wrongest company in the history of corporate lawsuits regarding its allegations, or Deel is hoping that it can simply keep operating sans expelling the leaders who thought that its spying campaign was precisely and exactly hunky-dory.
Deel’s board is presumably doing something, but I wonder if they are being cowards (putting paper returns over excising C-Suite rot from the company’s upper echelons), or are simply stuck (lack the voting power needed to oust allegedly morally-challenged leaders). Either way, not a great look?
Anthropic shows its teeth
Reuters has a great piece up discussing the growth that AI coding startups are currently bringing to market. Framed as a will these startups prevail over incumbent competition angle, the piece included a real zinger:
But [Cursor and Windsurf] operate with negative gross margins, meaning they spend more than they make, according to four investor sources familiar with their operations.
Negative gross margins do mean that the company is losing money, but more specifically, they indicate that AI coding startups are spending more on AI model costs than they generate in revenue for the same usage. That’s bad.”
And Windsurf is taking even more pain thanks to Anthropic, which has curtailed access to its Claude models and passed over it for Claude 4.0 access. From a March feature extolling the Windsurf (née Codeium) and Anthropic’s partnership to the AI model company “cutting off nearly all of [Windsurf’s] first-party capacity to the Claude 3.x models” that the smaller firm can access? What happened?
Two things:
OpenAI is widely reported to be in talks to buy Windsurf, meaning that Anthropic has little incentive to provide its technology to what could quickly become the fief of its arch-rival.
And Windsurf started releasing its own models last month. Again, Anthropic might not love that.
The present spat between the two companies is the first time that a model company cut a customer off, as far as I can recall. Expect sharper elbows in the coming quarters. Especially as Anthropic crests the $3 billion run rate mark.
Is tech buying slowing?
Looking at earnings from CrowdStrike (cybersecurity) and Asana (productivity) earnings this morning we see two healthy companies with solid trailing results. And guidance that left Wall Street wanting.
CrowdStrike turned in the expected $1.10 billion (+20%) worth of top line, and as CNBC notes eight cents per share worth of adjusted earnings per share than investors expected. But its shares are down around 7% this morning thanks to the company’s forward guidance of $1.144 billion to $1.152 foe the current quarter. Investors had penciled in $1.15 billion to $1.16 billion, depending on the analyst average you prefer.
That’s not a huge miss — and frankly after reading through the company’s earnings call transcript the trailing data seems far from lethal. The company is digesting some pricing changes that will lead to some near-term ARR weirdness, but those are small sums compared to its overall scale. Even more, CrowdStrike has a compelling AI story, even telling investors that it expects “improving sequential net new ARR growth next quarter and accelerating back half net new ARR.”
My read of the CrowdStrike situation is that investors are very skittish about anything other than full-throated confidence from companies regarding future growth, especially if they feature a market-leading revenue multiple. As CrowdStrike does today.
Shares of Asana are off an even sharper 11% this morning, despite the former startup turning in better than expected adjusted earnings per share ($0.05! Up from -$0-.06 last year!), and $187.3 million (+9%) worth of total revenue, above an expected $185.5 million.
So, why is Asana taking a pounding? It’s not because the company is dodging the AI wave. CEO Dustin Moskowitz told investors that his company’s “AI Studio, which reached general availability in Q1, surpassed [$]1,000,000 in ARR, demonstrating powerful early momentum” including “robust and rapidly growing pipeline” for the product in the second quarter.
Guiding to between 7% and 9% growth for the year ($775.0 million and $790.0 million), and 8% growth in the current quarter to revenues of $192.0 million to $194.0 million. The street had expected $193.0 million worth of revenue in the current period, and $781.3 million for the year.
So, what’s the problem? Demand concerns. Key quotes:
Anna Ramandi (COO): “[W]e are beginning to see some increased buyer scrutiny and elongation in decisions related to broader consolidation or software stack transformation efforts. That said, the pipeline remains healthy.”
Ramandi: “While the initial success has primarily centered on new business acquisition and expansion opportunities, AI Studio is becoming an important lever to mitigate downgrades in renewal conversations.”
Sonali Parekh (CFO): “[W]hile our in quarter NRR was stable for the third consecutive quarter, we expect net retention in Q2 to be pressured. This is due to a combination of continued downgrade pressure, particularly in our enterprise and middle market segments and the technology vertical.”
Parekh, discussing the company’s forecasts: “while we have not seen material change in the demand environment, we are observing early signs of increased buyer scrutiny and downgrade activity, particularly in our enterprise and corporate customer bases. While these early signs alone do not change our previous outlook, we do recognize that there is a growing macroeconomic risk and thus are reflecting this risk by expanding our guidance range.”
That last bit is the real kicker. Asana had guided between $782.0 million and $790.0 million for the year in its preceding earnings release. By lowering the lower-band of its guidance, the company is saying that current trends and added risk could be materially harmful to its growth. Asana reiterated that it is not yet seeing “additional risk and additional macro pressure,” but the company wouldn’t be saying those words out loud if there wasn’t some concern.
Down goes the stock. I would say that the above earnings reports are in many ways strong, and in a few ways slightly lackluster. The seemingly outsized investor reaction could be predicated on real concern that tech buying across key product categories (cybersecurity and business productivity, each with an AI twist), or merely worry about the economy at large with our two champions simply reporting at the wrong time.
But with a few other tech shops reporting guidance that has left traders wanting, we’d be remiss to not point out that — at the edges — tech buying could be slowing enough to matter.
Ceding the science baton
Closing, let’s talk about education and science. Something I’ve long been proud of as an American is our ability to attract global talent. If people vote with their feat — and tax receipts — then the United States is an amazingly popular nation. Myriad cultures brought together by a shared democratic-capitalistic foundation, it turns out, can generate some serious GDP.
But we’re working hard as a nation to cede that edge to others. Yes, the decision to make the United Sates inhospitable to foreign students is a mistake, as is applying thought-crime controls to national entry.
But past things that are politically fraught, we’re quite literally handing our arch geopolitical rival China a free win on the future. From The New York Times, discussing cuts to science budgets:
But with the Trump administration slashing spending on science, Dr. Patapoutian’s federal grant to develop new approaches to treating pain has been frozen. In late February, he posted on Bluesky that such cuts would damage biomedical research and prompt an exodus of talent from the United States. Within hours, he had an email from China, offering to move his lab to “any city, any university I want,” he said, with a guarantee of funding for the next 20 years.
Apparently yes, the CCP is busy over on the ‘Sky. And willing to spend like mad to get the future invented, designed, and built inside its borders. Alas.