I’ve been thinking recently about how to discover and hire great engineers in the hottest job market in decades. One of the biggest hurdles to hiring good engineers, and especially experienced engineers, is that they’re so. unbelievably. expensive.

Just take a look at some of the total compensation packages on levels.fyi:

Data is for GOOG (other companies are similar). Courtesy of the data at Levels.fyi (htps://www.levels.fyi/charts.html). This data was eyeballed quickly into a spreadsheet, check the source for actuals.

Total comp grows exponentially with years experience.1Granted, survivorship bias plays a part in this, but intuitively, it doesn’t explain the full extent of this, and certainly wouldn’t explain this trend if it holds up industry-wide and for new-hires coming in with X years experience. In some ways, this trend makes sense—even while demand for experienced engineering talent grows, that demand can only ever be filled by new engineers.

But this TCO graph also seems like a massive arbitrage opportunity: if I can give you two engineers with 1 year of experience under their belts, and they can out-perform a senior engineer with 10 years experience who would ordinarily make 3x their total comp, we can break this graph (I’m sure even as you’re reading this, you know this isn’t true…but this is essentially the argument that’s being made).

Both coding bootcamps and aggressive outsourcing are in part attempts to replace experience, pound for pound, with some sort of proxy (with bootcamps it’s careful vetting and exploiting the pay gap with other white-collar professions like CPAs or teachers; with outsourcing it’s exploiting cost-of-living disparities and the trust gap).

But rarely do these attempts to proxy experience work out. It seems like experience, at least in the burgeoning world of tech and companies with strategies that rely on digital excellence, is really hard to replicate or scale. And if anything the total comp curve is getting worse every year.

What gives?

The Googler’s Dilemma

I think part of the answer has to do with what I’m calling the Googler’s Dilemma, and it’s convinced me that not only is engineering experience irreplaceable, but if anything, the premiums companies are paying in total comp to these experienced engineers aren’t high enough.2Of course, years experience is itself only an imperfect proxy for true, applicable experience – something even harder to gauge – we’ll have to save that for a future post…

This is the Googler’s Dilemma: despite the fact that the internet’s accumulated experience is at your fingertips, it’s still inaccessible, because you don’t know what you don’t know, and therefore, can’t ask.

You don’t know what you don’t know, and therefore, can’t ask.

There’s a famous line people say about Los Angeles: “LA has everything…you just can’t get there!” That’s the Googler’s Dilemma. Thanks to Google search, there are plenty of proxies for experience available to everyone: you can read a few articles from top engineering leaders, stack overflow posts, gists, online courses, the list goes on. You can even read about other people’s mistakes and avoid those.

But all this information is available only if you know what to ask. And most people don’t. You can browse all the awesome-* lists and awesome-awesome-lists you want, but engineering experience is less about finding the matching answer to your question, or the perfect open source solution to your problem, and more about figuring out what question you’re dealing with in the first place. And that’s irreplaceable.

Some Counterarguments

Yes, it’s true—there are examples of massive unicorns that were built by kids straight out of college3Mark Zuckerberg’s lecture is a great piece of computing history, highly recommend watching who definitely were figuring it out along the way, no experience necessary. Sure, you could argue that if you’re experiencing hypergrowth, you’re contending with the reality that “systems survive one magnitude of growth” anyway.

Isn’t it better to optimize for smart people who just have a lot of grit to get through their mistakes, or who can move so fast that their mistakes fade into the background behind the blinding brightness of their intellect?

Absolutely. I have worked with some incredibly smart, talented engineers, and it’s game-changing for the engineering org. But banking your company on developing that sort of environment is risky as heck, and for most existing non-tech companies, it’s out of reach because you’ll be competing with FANG and there’s virtually no arbitrage left on hiring that profile of engineer. The only path available to them is hiring the right, experienced engineers, and that is a hard problem.

Here are some quick, high-signal hiring tips for finding experienced engineers that I’ve discovered over the years.