Things that don't scale: founder labor
14 / 10 / 2019 • Daniel Garnier-Moiroux
Recently, I read about a startup “success” story - in this case, like most startups, success means surviving long enough to bootstrap or raise capital. In the blog post that was published recounting the first few months, the inevitable hustle was mentionned - founders getting up at 4 in the morning to prepare orders for the day, doing everything themselves from business development to delivery. It vaguely reminded me of Paul Graham’s essay on “Things that don’t scale”. In a nutshell, the essay posits that it’s justified to do things that are not sustainable or don’t scale, for while. It is very likely necessary to do so if founders want to reach a “critical mass” were the ideas become a bootstrapped or growth-oriented business backed by investors. It is also most likely an unfair advantage you have over bigger competitors who have too many customers to do things that don’t scale.
The goal of many startup founders is to scale beyond the “small business” stage at some point. Even though you might not know the ins-an-outs of how exactly your business will scale, there are some assumptions you can make about how some processes will evolve. Founders fueling their startup with their own unpaid labor can only last for so long. The work still needs to be done, in an economically efficient way. What are the options when the business takes off ? Here are a few rough categories, which might overlap or be combined. Of course, most of them require upfront and/or continued investment.
Stop doing “the thing” entirely, or replace it with something different. In his essay, Paul Graham mentions founders sending a hand-written thank you note to every new customer. It’s likely that this won’t go on forever, if you have 1000 letters a day to write. Maybe founders will stop doing it entirely, or maybe they will replace it by a printed letter with a hand written “thank you”, not necessarily by them.
Make “the thing” extremely efficient through economies of scale. Think “building a factory”. When Peter Rahal founded RxBars, he would prepare the snacks in his mom’s kitchen, one by one, by hand. I do not know the full story of how they transitioned to a hundreds-of-million snack giant, but we can safely assume they invested in industrial production capabilities when they got their first sizeable investments.
Automate “the thing”. I’m a software engineer, a good portion of my energy is invested in automating things for myself, or for clients. There is often a tipping point when doing manual input in excel spreadsheets and sending them via e-mail is more costly than investing upfront to simplify that process - not necessarily all of it at once.
Delegate to your customers. Giving customers tools to do some of the work themselves - think self-checkout, booking stuff online rather than having to go on a call with customer service, etc.
Use cheap labor. Some things you’ll have a harder time automating, at least for now. Think: stuff that need interaction with the outside world, like delivery, things that computers can’t do very well (image labelling ?), or things where folks expect a human on the other hand (customer service ?). Given the way the world of startups currently work, this might end up in outsourcing or using gig work.
All of these have consequences - economic, social, environmental consequences. And while you can’t always predict how many pivots a startup will go through and how it will ultimately scale, it’s worth it to imagine a few scenarios ahead of time. You can then imagine if the consequences of scaling up are aligned with your concerns and values - or not.