Subscribe
Share
Share
Embed
The Diff is a newsletter exploring the technologies, companies, and trends that are making the future high-variance. Posts range from in-depth company profiles, applied financial theory, strategy breakdowns and macroeconomics.
How Companies Think About Layoffs
Layoffs.fyi shows the kind of growth trend you never want to see: over the course of 2021, they tracked a total of 15,000 or so tech layoffs. This year, the number is 151,648. To put this in perspective, every quarter starting with Q2 of this year had more tech layoffs than all of 2021—and last quarter's total was higher than the Covid-era peak in Q2 2020.
It's worth understanding the layoff process from start to finish: how do companies decide to do a layoff (as opposed to freezing hiring, cutting other costs, or letting attrition slowly shrink the workforce for them). And that also entails understanding the end goal. The bad news is that layoffs happen, and the structure of layoffs means that plenty of good people will lose good jobs. Sure, it’s frustrating to get laid off and immediately see your former employer hiring for a job suspiciously similar to the one you just got pink-slipped out of (I've been there!)—but companies that do a round of layoffs are generally restructuring for growth—so it’s locally unfair, but generally positive.
The High-Level View
It's helpful to start with the 30,000-foot view before we get to the "this plane is heading straight for a mountain and we need to veer off in a different direction" view.
In a functioning job market, there's inevitable friction that makes it impossible for any company to have exactly the right staffing at any moment in time. Finding the right person can take an unpredictable amount of time, and when the economy is humming along, people have a habit of switching. That's especially true if these employees are underpaid—i.e. the better a deal a company is getting on someone right now, the more likely that person is to leave. So the equilibrium is to slightly overhire. That's an especially easy choice for a growing company, because they can fix short-term mistakes through the bold decision to just slow things down a tad; a company that's 25% overstaffed but growing its revenue 100% annualized will solve the overstaffing issue organically in five months.
That friction also means that a healthy job market for workers is a nerve-wracking one for employers: wages are rising, and companies have to make hiring decisions in light of 1) the probability that many offers will get topped, and 2) the probability that some of their existing team will depart. Fortunately, this can be sustainable for a while when the perceived ceiling on employee productivity keeps rising. That's been increasingly true in software for a while; because when there's a SaaS company and an API for everything, far more scaling problems can be solved with money than ever before.
But this dynamic still relies on assumptions about the real world, and those assumptions won't always be borne out. Companies compete with peers on product quality and pricing, but they compete for talent and capital by being prudently optimistic. Whenever a big company does layoffs, it's common for people to say "Couldn't they have grown a little more slowly and never needed to do that?" And the answer is that there almost certainly was a competitor that grew more slowly and prudently, and that preference for stability over growth is why you haven't heard of them. In a space where every company is prudent and doesn't want to risk a layoff, the first company to defect and go for growth will swallow the entire market.
Outside of the industry cycle, there's a macro cycle, and that one is even harder for companies to avoid. Sometimes, their success is tied to real-world variables that they just don't have much control over. If a business operates in the travel space, there isn't much it can do about a pandemic; if a company sells electronics, it's probably counting directly or indirectly on China's manufacturing. For any B2B business, their revenue is part of somebody else's cost structure, so whether it's tied closely to transactions (e.g. payment processing, marketing) or indirectly to growth (almost everything else), a slowdown in corporate spending will have downstream effects.
When a company's revenue dashboard starts to persistently show that the numbers aren't going according to plan, decisions are required. That often involves diagnosing the problem, and figuring out why products aren't selling—is there new competition? A problem with an existing marketing channel? Or are customers just choosing to spend less?
If there's less demand than expected, the plans the company made around its original demand assumptions are obsolete. And that means that the last capital raise, which they thought was about growth, is retroactively the money they raised to ensure survival.
The Mechanism: Crossing Names Off Lists
The first dreary point to make is that the layoff process doesn't start with layoffs. It starts with more generic belt-tightening. Years ago, I worked at a company that replaced the fancy organic snacks with Lay's, and was confused by the grousing that ensued. You have a great job and you're complaining that the free junk food you're getting isn't as good as the free junk food that you're used to! But it turned out there was a reason behind the grumbling: companies don't like to cut costs, and the last thing they want to cut is lots of headcount all at once, so people who've spent years in big companies become attuned to this kind of thing.
Perks are, as it turns out, partly a financing vehicle. And like many financing methods that early-stage companies use, they can be quite expensive. The basic process goes like this:
A small company with negative cash flow knows that there's an exchange rate between dollars and survival. But not all dollars are created equal! Some kinds of spending can offset more than their cash-comp equivalent—especially if the early team a) has shared interests that make specific team outings or office features especially appealing to them, and b) have high enough morale (or big enough equity compensation) that they're not going to abuse an open-ended benefit like generous expense policies or unlimited time off.
As the company grows, the perceived value of these benefits to each recipient drops. But their cost rises at at least the rate of headcount growth—and probably higher, if anecdotes about employees waiting until catered dinner arrives and then taking home enough food to feed their entire family are to be believed. Meanwhile, the company starts to run into the problem that in-kind gifts tend to be value-destroying relative to cash. (It's seasonally appropriate to cite this paper estimating the deadweight loss of Christmas, here.)
But even when perks are no longer valued, getting rid of them sends a signal that the company either a) is suddenly worried about comparatively small sums of money, or b) is giving up on something that distinguished them from other companies. Both of these are bad for morale.
So generous perks early on are basically a way to borrow high morale from the future. Which can be a good deal! But it's still good to go into it knowing what the costs will eventually become.
When a company is growing and fundraising is easy, it's incredibly hard to resist the temptation to let costs creep up. But they can do so, in many different ways:
Hiring ahead of current needs means sometimes hiring for the wrong needs, and having people who won't be able to work on new priorities.
When a company is in a hurry and there's a simple subscription service that solves an immediate problem, it's very easy to just sign up for it and forget it. One customer requesting one fax one time years ago can lead to years of recurring payments for an online fax service; a temporary hack that uses Zapier to plug two or three services together can mean paying several vendors for something a single vendor can do better.
It's easy to bucket essential and inessential costs into the category of "customer acquisition costs that are worth paying because our customer lifetime value is so high." Corporate travel can really add up, in part because it's pseudo-lucrative: paying $2,000 to make a last-minute flight in order to have a 50% shot at closing a deal with $50k in recurring revenue is a very good deal, but it can be easy to over-attribute (and also easy to get in the habit of bleeding a few hundred dollars with every flight by waiting too long to schedule it).
Brand marketing fits a similar profile: it probably does have a high return in some cases, and for big companies that want to sustain growth the brand is key because it gives them a head start when launching new products—you probably won't buy a VR headset just because it's assembled by Hon Hai Precision Industry Co., but you might consider one from Apple.
All four of those ways are hard to measure, and they’re all focused on the long term. But when capital is less abundant, the long term is, financially, farther away, and risk is more asymmetric to the downside.
So companies start taking a second and third look at discretionary budgets, and start rearranging their product priorities: features that might attract new customers are out; features that retain existing customers, or that reduce internal costs, are in. (This is one reason downturns can be so reflexive: when there are fewer new customers to reach and companies are fighting harder to retain their existing customers, that hurts growth for their competitors, too.)
And, if everyone is flying coach, employees are brown-bagging lunch, the corporate holiday party is more downscale, hiring has been frozen for a while, and margins are still looking bad—well, that’s when the headcount reductions begin.
There are a few different schools of thought for how to choose who stays and who goes. But the general outline is:
Figure out what level of cost cuts would get the company to a more stable position. Stability depends on the business situation; it might mean having a path to cash flow breakeven, or it might mean having high enough margins to make the stock price tick up. (Or, for even more mature companies, it might mean being in a position to buy back lots of suddenly cheap shares.)
Look at specific projects that could be canceled completely. Sometimes this means shutting down something that just got acquired, or canceling an R&D project with returns that aren’t expected for years.
Meanwhile, make a list of unusually hard-to-replace people. This is a case where existing management has a serious advantage compared to, say, a private equity firm that's trying to clean house. Knowing that there's a system that only a handful of people understand, or knowing that there's a big customer who likes dealing with one specific salesperson, is the sort of local knowledge that's essential to reducing the negative impact of cuts.
Making cuts that get refined, from "% of payroll" down to individual names, as they make their way down the org chart. This sometimes happens at the level of managers cutting some share of the payroll expense of their direct reports, but can also happen a level or two up, when it's more driven by performance reviews.
Since firing is uncomfortable and hiring processes are imperfect, there's a set within most layoffs that's mutually beneficial: employees who weren't a fit for whatever reason, and would be more valuable somewhere else, get a severance package that will probably last longer than it will take them to get their next job. In other cases, it's a mutual disaster; a good performer who happened to get a bad performance review, or had bad luck with the team they joined, ends up losing a good job and moving to a less optimal one. The more a company has optimized effectively for fast growth, the more likely it is that laid off employees will fit this category.
Another relevant category is the people who were a great fit for the company when they were hired, but aren't that great of a fit any more. Some people do better in 20-person companies than in 2,000-person companies, and a layoff can be a forcing function for both sides to recognize that.
The Aftermath
It's tautological to say that companies do layoffs in order to cut costs. Of course a business that's spending too much money will need to reset its cost structure. But the important, and more interesting, question to ask is what end state they're aiming for. In some cases, the end state is that the company is going to switch from growth mode to just harvesting cash flows. This can involve pretty brutal headcount reductions of 50% or more. And it's a case where outside managers tend to do a better job, both because they can be unsentimental about people and because they aren't wedded to the idea of running a growth company. (Though there are cases where a successful reset like this comes from within.)
What’s more likely, though, is that the real goal is to reset to a more modest pace of growth, and for an environment in which there isn't a wild bidding war for every employee or sales/marketing channel that could conceivably add to the growth rate.
This is tied to another accurate piece of conventional wisdom: the best time to cut costs is before the situation is desperate, and the optimal number of rounds of layoffs in a given economic cycle is exactly one.
Put those together, and the optimal layoff is a lot more extreme than the change in growth trend that justifies it. A company might see that revenue is coming in 5% below expectations based on their current spending trajectory, might then deduce that the right growth rate is 10% slower than the previous plan, and might further conclude that it's better to aim for 15% lower growth and then accelerate from there than to risk a second or third round of cuts.
A good broad model of layoffs is that companies try to operate with various kinds of slack, and there are tradeoffs between them: a company that hires fast and pays generously can avoid asking anyone to work nights and weekends, but if the cost of this isn't justified by the revenue and funding situation, this increase in employee slack comes at the cost of a radical decrease in financial slack. Another source of slack is customers and suppliers—the same increase in discount rate driven by higher funding costs can be reflected in vendor renegotiations or in tightening up customer refund policies.
When the layoff process is over, if it's done right, the result is a company that's lean. And just like it's easier to maintain healthy habits—physical or intellectual—than to create them in the first place, it's now much easier for such a company to stay on track. It persevered through a difficult experience, and everyone who made it through wants to avoid more of the same, and, importantly, they don't have any of the classic bull market excuses for sloppiness.
Layoffs are not just a way to make the opex numbers trend in the same direction as revenue again. They're also an admission that this relationship got out of whack—that the pace at which a company added costs didn't align with its economic opportunities. That doesn’t mean that the company's broad thesis is no longer intact. Think about it like this: since the core theses of different companies varies a lot, if a company lays off 20% of its workforce, it implies that the hiring decisions it made were only 80% right, i.e. the workforce reduction is just a pace correction, not a pivot to a new destination. And once it's on the right track and has reset costs, the next round of hiring can be done more judiciously, too.
Layoffs are a financially and emotionally expensive form of time travel, a way for a company to go back in time to when it was leaner and more focused, and to rebuild from there. It's an admission of error, but the main point of admitting mistakes is to fix them.