Even if it doesn't fail and turns into a unicorn, its space needs will rapidly increase and it will still want to move on to bigger and better offices.
The entire point of a startup is to not stay its current size - either it will grow wildly or disappear altogether.
If you can comfortably occupy the same amount of space for 2+ years, you're not a startup, you're a small business.
Avoiding revenue does seem to be a good way to "grow" sans headcount growth, but that seems like a rare exception to the rule rather than a reason to believe that valuation growth doesn't generally correlate with employee count.
I was trying to make a subtler point with the counterexamples: that focusing on the wrong KPIs lead to unintended and occasionally negative consequences.
Quick example: imagine you are an IT consultant brought in to automate data manipulation processes in BigCorp (or more often MediumCorp since BigCorp has an internal team doing this already).
There's 150 people currently downloading data into Excel, running a few macros they know nothing about, and then reuploading the CSVs. You figure out what the data flow is, read up the API doc, and build some kind of process that does it in 5 minutes in bash on a medium instance on AWS, whilst fixing the errors the macros were making.
Will it be an easy sell? You were expressedly brought in to do this, but you'll find that the manager in charge of the 150 people - let's call him the CXO - is going to fight every inch of the way to stop you from launching your product. He'll point at the diff between his crappy but nevertheless, in production stuff and your script as cause for audit, creating a weeks/months long review process (because nobody can find the time, or wants to take the responsibility and the fight). He'll blame you for creating a "toxic" work environment. He'll bog you down in endless 2h long lunchtime meetings unrelated to the main point in an attempt to make you lose your professionalism in front of external stakeholders. He'll list missing new features then put them through the audit process.
Eventually, once a few months have been wasted, he'll point at your lack of progress as a sign of your incompetence, even though the working version was ready to be rolled out months ago. Nobody will question him because the company is profitable and it doesn't really matter how the process is done, just that it gets done. So, you leave the project, pocketing your pay - which they will pay on time so you don't make a fuss - and things continue as before for the CXO and his team, perhaps even gaining approval for a further 50 in headcount to develop the new features or to compensate for the expensive SQL and scripting training (MS stack, of course) that he's sending his entire team to in batches, whilst you'll have made a bad impression on that company's management and be burnt out of their network.
The CXO benefits outside the company as well, as he can now say he managed XYZ employees which is how people automatically gauge someone's success. This enables a succession of ever greater responsibility positions (or entry to Harvard Business School, who actually asks you "how many reports did you have"). "Well, if the previous place trusted him with 150 people..."
It's even worse in startup space because VCs and many founders understand that growth at all costs is what matters, so there are significant incentives to doing it today, even if badly, vs doing it tomorrow but well and in a way that doesn't pile technical debt. It's a winner takes all, so you just need to pour in enough millions and you'll reap the billions (in fact this is also the very structure of a VC's portfolio, investing in 10 different ways of doing the same thing in the hope one wins, even if it means the others all die).
So some founders may be doubly incentivized: making the company as large and cash flow burning as possible both to appear like the obvious winner, and to justify mammoth fund raises at a time of abundant capital seeking yield; and, if things go wrong, well, they managed hundreds of millions of VC dollars and hundreds or thousands of people in dozens of global offices and it sounds damn good for their next try.
And this is where both WhatsApp and Instagram did things differently: they focused on getting the user growth scalable without enormous headcount growth, whilst - at least in Whatsapp's case - having enough cash flow to hold on until the mega acquisition (and in Instagram's case, sustaining the company on the early rounds). Pointing at lack of revenue is not particularly useful - I would say even a red herring - when even SaaS companies go 5-10 years without showing profitability and with enormous fund raises for growth (just look at the analytics space for recent examples).
The hedge fund space - which is almost by definition results driven - has already caught on - Bridgewater type funds with a thousand employees are the exception rather than the rule and seeing AUM over a billion USD per head is more common than "unicorn" startups.
Growth in users or revenue does not always cause growth of headcount, but they are definitely correlated, especially in organizations with sales teams.
Of course, but I was countering the OP's point that "If you can comfortably occupy the same amount of space for 2+ years, you're not a startup, you're a small business."
I agree with pg's definition of a startup as being about growth (in the eponymous essay). In users, not headcount, in revenue, not funding. Even if there are many cases where one drives the other, targeting headcount growth and fundraise amounts as success KPIs is an instance of the https://en.wikipedia.org/wiki/Cobra_effect.
"A startup is a company designed to grow fast. Being newly founded does not in itself make a company a startup. Nor is it necessary for a startup to work on technology, or take venture funding, or have some sort of "exit." The only essential thing is growth. Everything else we associate with startups follows from growth." - pg
I guess everyone is entitled to interpret words however they like, but the salient attribute of a startup is exactly what the word means: it's a new venture. And from the perspective of a landlord, that's what matters. New companies are the least predictable in their needs, their revenue stream, and their long-term prospects. They tend to have inexperienced managers and flighty owners. All of these things are true regardless of how they are funded, what industry they operate in or whether the owners are trying to become huge overnight, bootstrap into a comfortable living, or whatever else.
Established companies grow and shrink, too. They close offices and factories, open new ones, and occasionally file for bankruptcy protection. But they do these things much less frequently than a new company, and they do it with the backing of a much larger and more reliable cash flow. That cash flow means that even in bankruptcy, an established company's creditors usually end up with a significant recovery. From a landlord's perspective, these are the things that make established companies more attractive than startups. Thus, in this context at least, "startup" means nothing but "new company" and carries none of the other connotations that most people here assume when hearing the term.
There are many other types of business besides startups, and not all technology businesses are startups. That's fine, many of them work very well. Rapid growth is not the only way to run a successful business. But they're not "startups," by definition, they're some other kind of business.
You can, of course, attempt to appropriate the word "startup," redefine it to mean something else, and then attempt to get others to use your new definition, but empirically it seems that almost everyone is using pg's definition above for that word.
The entire point of a startup is to not stay its current size - either it will grow wildly or disappear altogether.
If you can comfortably occupy the same amount of space for 2+ years, you're not a startup, you're a small business.