3 ms·
I agree with this take, and I'd also add: the broader macroeconomic issue with SaaS concerns how quickly companies lose its utility from downsizing their expens
by 015a 4y ago
I agree with this take, and I'd also add: the broader macroeconomic issue with SaaS concerns how quickly companies lose its utility from downsizing their expenses. If I, as a customer of Canva, stop paying for Canva; its safe to assume I was paying for it for a utilitarian, revenue-generating reason, and if I cancel it I immediately lose the ability to capitalize on that utility. This hurts my business, which then cascades further. In comparison, choosing not to upgrade to Office 2007 or whatever isn't an ideal choice, its one a healthy business may not make, but it is one that an unhealthy business attempting to survive a period of lowered revenue may make; Office 2003 works just fine.
In that sense, the cascade effect is actually bidirectional; canceling Canva immediately hurts Canva, but it also hurts my customers; more acutely than, say, waiting a year to upgrade to the latest version.
Additionally, many people talk about bullshit jobs, and it seems obvious to me that subscription based billing has been a driving factor in this phenomena. If, revisiting the previous example, Office 2007 doesn't deliver enough value to Me The Customer to convince me to upgrade; I won't. Under subscription based billing, I'm given no choice; it follows that we enter a pattern where companies are far less incentivized to deliver innovation. When speaking on a domain where there's little competition, this effect is heightened, but the move to subscription-based billing subtly removes even another competitor from the domain: your past product.
The other thing I think about when it comes to software companies: how little inherent value they have. While not the only area where it matters: think about bankruptcy liquidation. Their value (at insane valuations, to be clear) was in the service they provide, but when the rubber hits the road the code really doesn't have much value (its kind of like that old story about the guy who stole coke's secret formula and tried selling to pepsi; even if you could obtain it legally, why would you want it?). Maybe the data does, for some companies. Most of these companies don't have servers; they may buy their employees laptops. Many are moving to full-remote; so no real estate (which may have been rented anyway). When revenue starts plateauing and the veil is lifted, the reckoning will be so much harder.
Point being, I think the pricing, expense, and operational model in typical SV-style corporations is subtly unsustainable.