When I was twenty-two, an older engineer in Lagos told me: pick one thing and become unkillable at it. He had built his career on a single database. He could be flown into a room and paid a great deal of money to look at one kind of problem. He was, by any reasonable definition, winning.
I tried, for about a year, to take his advice literally. It almost ruined me.
The advice that doesn't fully translate
Most career writing about specialization is written from inside economies where the market for narrow expertise is already large enough to absorb a narrow expert. Become the best Postgres performance consultant in your city, and there is a city with enough Postgres workloads to pay you for it. Go deep on payments infrastructure, and there are fifty companies in a 50km radius that need exactly that.
Nigeria does not work that way for most specializations. Not yet. The Postgres consultant in Lagos has, on a good day, four companies that need her. On a bad month, she has zero, and she is competing on price with a generalist from Ibadan who will also do the React work and the DevOps and the Slack support, all for the same fee.
This is not a complaint about the market. It's the actual shape of the market, and career advice that ignores the shape of your market is just expensive cosplay.
What I did instead
I became, deliberately, a narrow generalist. The phrasing is borrowed, but the practice is local:
I picked one thing to be genuinely deep in — backend systems under load — and I built a wide, cheap-to-maintain periphery of skills around it: enough frontend to ship a product alone, enough infra to deploy and debug it, enough product sense to argue with a founder, enough writing to make a case in a Slack thread.
The deep skill is what made me hireable for serious work. The wide skills are what made me employable at all in months when serious work wasn't available — which, in Nigeria, is most months for most engineers.
A specialist in this economy is a generalist who has earned the right to be paid for one thing, while quietly doing four.
The mistake I see often
The mistake I see most often, in juniors who read the same career books I read, is to chase American-style specialization in a market that won't reward it for another decade. They turn down the contract work that would have built their portfolio, because it isn't on-spec. They refuse to learn the second language their employer needs, because they're specialists. They optimize for a career that, structurally, doesn't yet exist for them where they live.
The other mistake is the opposite: an indiscriminate generalism that has no center. The engineer who does a little of everything and nothing memorable. This person is also losing — they are competing on price with every other generalist, and the market for general-purpose engineers in Lagos clears at a number that will not pay rent in five years.
The point isn't to refuse specialization. It's to know what you're specializing into, and what you're specializing around.
A small heuristic
When I'm advising someone three or four years in, I ask two questions:
1. If a serious problem comes through a serious door, what is the one thing you want to be in the room for? 2. If that door doesn't open this quarter, what are the four things you can credibly bill for in the meantime?
The first answer is your specialty. The second is your runway. In Nigeria, both matter. In a different economy, maybe only the first does. We don't yet live in that economy.
The older engineer who told me to pick one thing did, eventually, leave Lagos. There were not enough Postgres rooms here to keep him fed. He works in Berlin now and is, by all accounts, doing fine. His advice was right for the economy he ended up in. It was not, exactly, advice for the one he gave it in.