What We Talk About When We Talk About Hiring Talent in Tech

Sean Owen
5 min readAug 4, 2015

--

“Hey, do you know any great tech people looking for an opportunity? I’m hiring.”

Like junkies shaking down their Rolodexes for leads on more skag of a Friday night, anyone running a team or company in tech seems to be endlessly asking each other this.

“All the good folks I know are happily employed, sorry!”

… we all tell each other and then promise to share any leads and shuffle on. Always Be Hiring! Why does it seem so desperately difficult?

My kingdom — or at least, a small percentage of equity — for a unicorn!

I sat down and collected some observations, mantras and received wisdom I keep hearing about working in tech, and hiring great tech folk, in order to figure this one out for all of us. What I found may surprise you: it doesn’t make sense. Consider the following propositions:

  1. Tech recruiter e-mails are so lame and ineffective
  2. We get a bunch of lame unsolicited CVs
  3. We need experts in the following N cool new technologies
    (or: “data science”)
  4. Join us: here you get to work with N cool new technologies
    (or: “data science”)
  5. We only hire the best of the best
  6. We pay market rate compensation and don’t give raises
  7. There’s a tech talent shortage

Individually, I dare say these are recognizable as things people say, a lot. But most groups don’t make sense side-by-side.

On Market Rate

5 + 6 + 7 is the most glaring and fun one, so let’s take it first. As The Register says, it’s like complaining that there’s a Ferrari shortage. In fact, it’s like complaining that there is a shortage of Ferraris at Ford Focus prices.

Reddit’s erstwhile CEO Ellen Pao even suggested they view negotiating salary as simply not done because it might advantage people who negotiate, like, men on average.

Talk of sticking to “market rate” sounds a little like collusion to not compete on price, so, let’s not say that. Because the tech labor market does not collude.

There are good, but ultimately orthogonal, arguments against big pay discrepancies: it rewards slick negotiators, it creates internal resentment, etc. But why is this so off the table?

On Tech Recruiters and CVs

1 + 2: Lots of people selling money-for-work opportunities, or jobs; lots of people seeking money for work. Can’t these people just talk directly to those people and leave me out of it? Why is this a thing?

1 + 5: If there’s no willingness to pay extra for top talent, a rational response is to find under-priced talent. Cynically then, recruiters are playing a numbers game to find these rare mispricings. The actual outcome is usually just normal talent for normal pay. But if great talent is hired this way, the talent must be being seriously underpaid right?

But nobody apparently bats an eye at acqui-hiring good engineers to the tune of $500K to $1M or so. It’s somehow not about the money: companies will pay plenty for talent, sometimes. Talent doesn’t, apparently always care about money. Wha?

2 + 7: A talent shortage, but, so much talent knocking on the door. Yes, but, that software engineer I bothered to phone screen said he had 10 years of distributed systems experience but though Paxos was an island! any distributed systems engineer should know what a consensus protocol is. Why would someone like that bother applying?

On Responsibility As Reward

This brings us to 2 + 3: We put out a requisition for “data scientist” because we need a data scientist. This woman who applied is obviously a data engineer — not even any Python experience, ugh! Why would someone like that bother applying? Data scientist = Python person!

Reqs for something like “software engineer” are almost by definition hopelessly broad. It’s like hiring an “athlete”. Just, an “athlete”. Reqs may be born from a real need. We need someone to take over these Python scripts and productionize our video recommendation process from it. But even in a no-nonsense small startup, the reqs turn into wish lists: Python experience yes; let’s say lots of it. R would be nice too. Is an advanced degree in math important? probably, put that down. Candidate should know Hadoop, Hive, Spark too and have a winning smile. And now adjust that for a big company filtering through a shared recruiting team. You get things like IBM hiring for 8 years of Cloudera experience, 7 years after its founding.

They become reqs for a unicorn, or nonsensical, and by asking for most everything they ask for nothing, and suddenly people who just fit some part of the bill apply. Why do we do this to ourselves?

3 + 4: Because, if we actually describe what it is we need done, it might sound boring or it might pigeonhole us. We can’t literally say we just need a junior Python person who’s willing to learn, right? Nobody wants to just do Python, right? So, a job requirement is actually the opposite, a selling point: when we say “6 years Apache Spark experience required,” we mean, “you get to work with Apache Spark!”

Here is the Answer

Just kidding — I have no idea what the answer to all this is.

But, then I remembered a post by a smart person at a well-known tech company, which sticks out in my mind so much that now I can’t find to save my life. (TODO: insert link when someone reminds me what I’m talking about.) It has, I think, part of the answer.

The gist was, sort of: attract talent by offering honest career growth. The conversation should turn around why a role would get someone from to the next career step on the scale of 4–5 years. It’s learning and professional network that are on offer in return for doing great work for market-rate compensation. The company does need someone to get a job done, and not merely contribute to the awesome culture, rate the kitchen’s artisanal froyo, and be deeply pumped about a shared mission to gamify a pet dating site.

So, maybe that’s why we can go back to accepting that market-rate salaries are fine for everyone. As employers, we need to be specific and clear about what we need to get done and what skills that definitely requires. But equally we need to advertise what else you can master in a stint here — what you can become famous for while here if you want to. Don’t hunt desperately for the unicorn or outsource to unicorn hunters; be a tech unicorn training ground.

Candidates: be clear about what you can do and what you want to learn to do by the time you’re 4–5 years further on. Find a role that can reward your talent with growth rather than focus on money or title. The best talent can certainly demand excellent growth prospects.

As far as I know this is the answer.

By the way, do you know any great tech people looking for an opportunity?

--

--

Sean Owen

Big-data data science personality @ Databricks. Prev: Director Data Science @ Cloudera