What Taylor Swift Can Teach Us About Enterprise Software Engineering
Taylor Swift is one of the most commercially successful musicians in history. This has almost nothing to do with software engineering, which is precisely why I am writing about her.
I like writing about whatever happens to interest me, which has produced essays about subjects ranging from Old Dirty Bastard’s FBI file to the Ballmer Peak to the relative merits of caffeine pills and coffee. Our marketing team made an interesting observation about this: every time we publish something, we’re providing Google with another piece of information about what Gun.io is actually about.
They gave an example that we could probably write an article about Taylor Swift and generate traffic. But the question is: would it bring the right readers to our site? This is a perfectly reasonable point, and our marketers are right.
It also immediately made me want to write an article about Taylor Swift.
Fortunately, there actually is an interesting connection.
Taylor Swift is an unusually obvious demonstration that talent is not evenly distributed. You could assemble a very large number of perfectly competent singers and songwriters and still not reproduce the economic output of one Taylor Swift. This is obvious in entertainment because we don’t expect talent to scale linearly. Nobody thinks that ten musicians of one-tenth the commercial value can simply be combined into Taylor Swift.
But for some reason, businesses have historically been more willing to make that assumption about knowledge workers. If we need twice as much productive capacity, the intuitive answer is often twice as many people. Headcount becomes a rough proxy for output, and organizational growth becomes closely associated with hiring.
Software engineering has never worked perfectly that way because differences in individual ability can be enormous. But technology is making the distinction increasingly important. As AI increases the amount of work that a capable engineer can produce, the relationship between headcount and output becomes even less linear. The important question increasingly isn’t how many engineers you have. It’s what the engineers you have are capable of accomplishing with the tools available to them.
This doesn’t mean every company should replace ten engineers with one superstar. Large systems still require teams, specialization, redundancy, institutional knowledge, and people doing work that cannot simply be multiplied through tooling. But it does mean that adding people is only one way to add productive capacity, and increasingly headcount additions may not be the most efficient one.
That has interesting implications for how companies think about technical talent. If technology increases the leverage available to an individual, then differences in talent matter more, not less. A highly capable engineer equipped with increasingly capable tools can operate across a larger surface area, which makes finding the right person potentially more valuable than simply finding more people.
At Gun.io, this is a subject we happen to spend a lot of time thinking about.
And there it is. Our marketers made a perfectly sensible point about topical authority. I wrote the Taylor Swift article anyway. And, through a sufficiently determined act of editorial engineering, it is now genuinely about software engineering.
Reader, please make a note of that.
