I currently have two jobs.

During the day, I lead a merry band of engineers writing an Angular app that enables developers at my company to get DevOps pipelines set up on their various projects easier.

By night, I make web sites, by myself, for clients of the Tiger Pajamas Web Site Company.

To an outside observer, even a technically minded one, these jobs look very similar. I am writing HTML, CSS, and JavaScript(1) into Visual Studio Code, committing and pushing the changes to a git repo, and those changes are built on a remote server and served to end users through a web browser. You could call both of them “software engineering”, or “front-end development” if you want to get more specific. You could, but you would be wrong to do so.

Tiger Pajamas is very specifically and deliberately called a “Web Site Company”. We are not Tiger Pajamas Software Services or Tiger Pajamas Technical Solutions. What we make is on the tin. Web Sites. And it’s worth making the distinction here, because it turns out that almost everything about our business is so alien from what a “software company” does that we are basically speaking different languages.

What is software

Some basic economic history would be useful here. Software is a relatively new profession, emerging in the late Industrial Age and defining the Information Age that we now find ourselves in. Its industrial roots are visible in linguistic vestiges: “production” is the word we use for live servers, and it shares a name with the factory line that produced finished goods. It also had more earthy roots, back when programming a computer meant physically plugging and unplugging wires between vacuum tubes, or soldering chips to a circuit board. As the field has matured, software has grown more abstract, and its distribution model started to differ wildly from the hardware it inhabited. You write a computer program once, and physical media like floppy disks and CDs made it relatively cheap to meet the market with as much supply as required. Now, distribution is effectively free, meaning there is potentially tremendous upside to making something that can scale up to hundreds, thousands, millions of users.

Finding a common source of business inefficiency and creating software to address it is a great way to make money. A platform that helps reduce something like supply chain inefficiencies can price itself out of the cost savings it provides, and provide a stable living for the people who work on it. Or, a social network that helps undergrads hook up can become the ubiquitous town square for over half of the world’s population and make trillions of dollars by monopolizing online advertising. Y’know, just as an example. The sky’s the limit!

The whole thing is driven by market-based capitalism, that ubiquitous system we all know and love. Infinite scalability is a natural counterpart for infinite growth, the goal of the monstrous corps-tumors whose shadows we toil in. Shareholders love when line go up, and the various apps that have taken over the entire world are very good at line-tilting! And so software that serves these goals tends toward very specific kinds of shittiness: user interfaces that are, to use a technical term of art, “wack”, threadbare documentation and support, and dark patterns that trick our lizard brains into pouring more and more engagement and money into them.

This sucks and is bad. Not to say that all software is like this, but the market forces that pay for it relentlessly bend trajectories in this direction, and it takes tremendous and persistent effort to resist it.

What is a (Tiger Pajamas) Web Site?

Let’s use a metaphor. Software is the $10 sheet cake that you can buy at the grocery store. It’s a bog standard product, available within a 10 mile radius at a moment’s notice, and they’ll even scrawl “Happy Birthday ${BIRTHDAY_BOY_NAME}!” on it for free. Customization!

A Tiger Pajamas Web Site, on the other hand, is more like a wedding cake at the kind of wedding that makes it into the Sunday Styles section of the New York Times. You starting to see what I’m getting at here?

The incentives are so different that it’s useless to even try to compare them. So useless that it’s actually going to confuse you more if you try. You know what I spend most of my time writing? CSS! The vast majority of the labor required to make one of our sites is the work Ash does to create a unique, personalized, soul-bound design, and for me to execute it. Of course, JavaScript bros will tell you that CSS is a waste of time. I don’t go more than a week without hearing a talented CSS developer bemoan the fact that their skills aren’t valued, and it’s true, because it’s very hard to convince line tilters that pixel-perfect positioning, sitewide accessibility, and layouts that relax and invite the human eye “drive value”(2). In the world of software, the beating heart of value is the business logic that reduces the inefficiency, everything else is an annoying chore.

For us, it’s almost the opposite. We are, of course, experts at the modest efficiencies needed to deliver sites that load fast and don’t waste bandwidth, but once those pesky details are out of the way, everything else is about giving our clients a friendly, intuitive, and fun space to inhabit, to enhance the joy that comes from creating something special, and for their users, the joy that comes from finding something online that speaks directly to you. You can try to squeeze that into an efficiency shape if you want, but I’d recommend not doing so, because you will kill it dead and then you’ll be left with nothing. The human soul flourishes when it is free of the need to be Efficient, Optimized, and Busy.

What remains to be seen, as we do this work part-time around our day jobs and childcare and household responsibilities, is if we can make this business our entire livelihood. That’s a big question, but I’m optimistic. Because market forces point so strongly in one direction, we’ve discovered vast stores of pent-up demand for skilled web craftsmanship, and our unique experience is very potent. We do not compete as a commodity; you cannot get a web site sold by Jon Mitchell, designed by Ashley McQuaid, and coded by Phil Giammattei anywhere else, so we have wide berth to set our prices. We have not tried particularly hard to find work since our first meeting in March, and already we have shipped two sites, have 2 more on the way, and then we might have to put a hold on new clients so we can actually work on our own site!

The way is long and hard, but it’s not like we need to capture “market share” or whatever. There is no scaling; we are never going to outsource or hire because the whole point of the work is that we’re doing it. We are free to spend as much time as we want to polish our work to a mirror finish, without compromising our values. It feels very good to do this.

It’s easier to find a job as a sheet cake baker in a grocery store bakery, but if you put in the work, you might get paid far over market rate to make a cake so well-crafted that it looks like something that isn’t a cake, and then they put your ass on TV, baby.

1: Mostly TypeScript now, actually, but in the interest of simplicity…

2: Of course, the whole thing smacks of gender as well, but we’ll save that one for its own post.