Google’s postpandemic ‘reckoning’

This week, Shailesh Prakash, the VP in charge of Google News, gathered his hundreds of employees to talk about a surprise round of layoffs that had just hit the team.

A year ago, cuts like this at Google — historically the most coddling employer in all of Big Tech — would have been unthinkable. But market forces have a way of catching up with everyone. In January, the search giant conducted its first mass layoff in over a decade, cutting 6 percent of its staff, or about 12,000 people. Last month, hundreds of recruiters were let go, a concrete sign that hiring targets have been reduced significantly.

The layoffs this week in Google News were relatively small, with roughly three dozen people impacted. But why they happened has bigger implications, especially for those who are more senior across the sprawling conglomerate that is Google and Alphabet. At his all-hands meeting for the Google News team on Wednesday, Prakash said the company had grown headcount too quickly during the pandemic and that “there’s a reckoning” being felt now as a result. According to people who heard the remarks, he specifically called out the influx of level 8 and 9 roles — some of the most senior positions one can hold before becoming an exec — that Google either hired for or promoted people to over the past couple of years.

Prakash also echoed the explanation for the cuts given by Google PR to CNBC, which is that they were intended to “streamline” the org. Unlike the rushed, seemingly random mass layoffs in January, Prakash’s comments indicate that a new, more surgical phase of Google’s postpandemic reckoning is underway. The News division cuts were quite targeted, I’m told, with several “director decapitations.” Additionally, I’m hearing that multiple VPs across Google have yet to reach their headcount reduction targets for the year. Altogether, this suggests that a slow drip of cuts targeting senior folks could continue.

Privately, managers at Google will admit that the company is still far too bloated. There are plenty of anecdotes you’ll hear of people making ungodly amounts of money to do duplicative work or barely work at all. (L8s and L9s at Google are making over $1 million a year.)

You’ll also hear an increasingly sour view of upper management and how they have handled cost cutting this year. By continuing to prolong the fear and pain, Google is at risk of burning a lot of goodwill with future recruits that it spent many years building, one current manager told me this week. Sometimes, it’s better to just rip the Band-Aid off and move forward.

Congratulations to those at OpenAI who are able to cash out in this new tender offer, which is happening at an astronomical valuation of more than $80 billion, or more than 40 times this year’s annualized revenue. Thrive Capital, which led the last buyback in April at a $27-billion valuation, is leading this one, too.

Unlike a traditional funding round, these buybacks are really for employee retention and recruiting. If you’re a sought-after AI researcher evaluating offers from Big Tech and the other AI labs, it’s hard to pass on the liquidity that OpenAI provides (assuming the valuation continues to climb, of course). A lesser-understood aspect of OpenAI’s compensation is that employees are given profit participating units (PPUs) rather than the traditional stock options or restricted stock units (RSUs) you normally see given by tech companies.

Unlike stock, which gives you the right to own a part of the company, these PPUs entitle OpenAI employees to a capped percentage of the company’s profits, should profitability ever arrive. I’ve heard new recruits are told that day likely won’t come until OpenAI archives AGI, or artificial general intelligence, which leadership thinks may happen in the next decade.

Compared to RSUs, PPUs are heavily tax-advantaged. And you don’t have to buy them when they vest like stock options. By allowing its PPUs to be traded like stock, with outside investors regularly buying them from employees through tender offers like this one, OpenAI is exploiting a clever loophole that has made it perhaps the most lucrative place to work on AI right now, even as it burns billions to build its expensive models. Who needs profit when you have VCs, right?

Speaking of buybacks, ByteDance, the parent company of TikTok, has shown how not to do them. I’m told that roughly two dozen complaints have been submitted by former employees to the SEC after I recently reported on mounting frustration at how ByteDance has gone about the process.

Beyond the perplexing decision to offer former employees a lower share price than current employees, many are shocked at the surprise tax burden they now face for finally being able to sell the stock they’ve vested over years. Thanks to the way RSUs are taxed, and the fact that ByteDance is just now offering liquidity to US employees for the first time, people are getting kicked into higher tax brackets unexpectedly due to their stock vesting all at once, even if they decide to not sell back their shares.

For those who recently opted to receive RSUs over a cash bonus, the sting is especially strong. Many were granted this stock at $158 a share, which means they are being taxed at that price while also losing money at the $128 price that ByteDance is offering former employees in this buyback. Ouch.

Communication I’ve seen from ByteDance to angry employees can be summed up as deal with it, which isn’t sitting well. It’s certainly not the company’s fault that employees are being surprised by the tax implications. But it’s increasingly looking like this episode will end in ByteDance looking like a much less attractive place to work.

Thanks to all of you who wrote in and subscribed after last week’s issue on Snap, which seems to have struck a nerve inside and around the company. After I reported on the business goals for next year, the stock price went up more than 10 percent.

Bernstein analyst Mark Schilsky told Reuters that, while expectations for the stock are “incredibly low, even I have to admit that the stock would likely be materially higher” if the goals I reported on were hit. “It’s hard to find a more hated company” in tech, a large Snap shareholder told me this week, though this person was quick to note that investors shouldn’t discount the upside that still exists if the business turns around. Very few platforms ever reach a billion users, and Snap has a path to achieving that milestone over the next few years on its current trajectory.

I’m all for measured optimism. But I was a bit surprised by the positive reaction to the stock after my report. Sure, Snap hoping to hit 20 percent revenue growth next year is above what Wall Street was expecting. However, this is an internal stretch goal. If you were to grade Snap’s stretch goal achievement over the past few years, it would be an F. In 2018, a stretch goal was to be full-year profitable in 2019, for example.

We’ll know more about the state of Snap’s business when it reports third quarter earnings next week. In the meantime, here are a few reactions to last week’s issue that I rounded up:

“Underneath the diplomatic tone, it’s a pretty damning - but at the same time objective - take on Snap’s leadership, operations and prospects.” – Tidal Capital

“Snap’s in a pretty tough spot, Apple’s ATT hindered ad growth and the company has not been able to adjust as quickly as Meta. [Evan] Spiegel trying to ramp up Snap+ subs quickly suggests the company will double-down on Al lenses/bots to boost revenue, but when everyone else is offering these features for free, how do you convince users to keep paying $3.99/month?” – Trevor Croker, editor at Techmeme

“Same challenges as always - Snap is a really good product, but the company has never developed an actual business around it. Is 12 years old with over $1B/quarter in revenue, but is still quite unprofitable. Market doesn’t have the patience for this type of story anymore.” – Sean Ryan, a former Facebook VP

“I have no interest in trying to compete with the smartphone. It’s phenomenal at what it does.” – OpenAI CEO Sam Altman

I was curious to hear Altman address the rumors about his interest in building a hardware device with Jony Ive during his appearance at The Wall Street Journal conference I attended earlier this week. He downplayed the idea that he would make anything resembling the “iPhone for AI” and made it sound like any talks about OpenAI making hardware, including its own AI chips, were very nascent.

Regardless, I’m of the opinion that OpenAI needs more distribution, whether that be a consumer device or linkup with a large internet platform, as the generative AI race heats up. Unfortunately, I didn’t get to hear Altman’s take on this topic this week at WSJ’s conference, though I did spot him deep in conversation with Meta CPO Chris Cox at the hotel bar late into the night…

A roundup of what else happened in the tech industry this week that you may have missed:

That’s it for this week.

I’ll be back next week. In the meantime, send me your feedback, story ideas, and writing tips for Marc Andreessen. Thanks for subscribing.