How we scaled a Philippine team from 12 to 240 in 18 months

Eighteen months. Twelve people to two hundred and forty. The head of operations walks through the actual mechanics — not the highlight reel, the parts that hurt, the assumptions that broke, and the two or three moves in retrospect that made the difference between healthy scale and the version where everybody’s burned out by month twelve.

The context matters: consumer product, Series B, Philippine ops centre, an existing 12-person team that was already stretched. The mandate was to hit 200 within a year. They hit 240 in eighteen months. This is what actually happened.

The operating model, in three moves

The first thing we did was split the recruiting funnel by role archetype, not by seniority. Trying to interview a customer-success hire, an infra engineer, and a growth marketer through the same funnel wasted everyone’s time and delivered mediocre candidates in all three tracks. Separate funnels, separate rubrics, separate hiring managers who owned end-to-end. This alone took our time-to-hire from 42 days to 19.

Second, we changed how internal promotions worked. No one got promoted to a management role until they had trained the person who would take their previous seat. No exceptions, no “acting” titles, no promises. This was unpopular in month one and load-bearing by month six. It forced a real conversation about succession — which is the same conversation as “what does good look like in this role, actually?” — and it starved the org of the kind of manager who gets promoted for tenure alone.

Third, we ran a monthly friction audit. Any team member could nominate a process for deletion. Anonymous, low friction, hosted in a shared doc. If a nomination survived 60 days on the list without a senior leader publicly defending why it existed, it died. In the first year we killed 34 processes. Six of them, we probably shouldn’t have. The rest were exactly the accumulated crap that quietly kills velocity in a fast-scaling org.

What broke, and where

The ceiling was manager quality. Not hiring pipeline, not compensation, not tooling — those were solvable with money and time. Manager quality was a slower problem, and it hit us at exactly the moment we were least equipped to notice it, which was somewhere around 80 people.

The classic symptom: your best ICs stop getting time with their manager, because their manager is now spending 40% of their week doing performance conversations with people they were peers with six months ago. Your good ICs start looking around. You lose two of them and don’t understand why, because their performance reviews were glowing. You lose two more and start understanding. By the time you understand, you’ve replaced a third of the layer beneath the manager and burnt eighteen months of context.

The fix wasn’t more manager training. Manager training is a slow-acting intervention against a fast-moving problem. The fix was fewer managers with wider spans, and better ICs beneath them. We flattened. We took the median span of control from 5 to 9. We paid the strongest ICs on the individual track more than the middling managers we would otherwise have kept. Counterintuitive, expensive on paper, and it worked.

The single biggest lesson: at ~80 people, the temptation is to add management. What the org actually needs, almost always, is subtraction.

What we would do differently at 240 → 500

Three things, in decreasing order of certainty:

1. Hire a Head of People before the CEO thinks they need one. We had ours nine months later than we should have. In hindsight, that’s the difference between the version of the org that developed a healthy performance-management culture from the inside, and the version we actually got — which was one of us duct-taping performance reviews together on a spreadsheet at midnight and hoping nobody noticed the inconsistency.
2. Build an internal-mobility function on day one, not day 500. The best candidate for your next senior role is almost always sitting three seats over already. If you don’t have an obvious path for them to raise their hand — and an unbiased way to evaluate them against external candidates — you’ll hire outside and lose them.
3. Instrument everything, but publish only what people can act on. We had beautiful dashboards nobody read. What actually moved behaviour was a weekly, three-metric email — retention, offer-accept rate, and one focus metric per department that changed quarterly. That was it. Everything else was noise dressed up as data.

The parts that don’t make the highlight reel

You’ll fire someone you shouldn’t have. You’ll keep someone six months too long. You’ll under-comp a person whose loyalty you will later spend a lot of money trying to buy back. You’ll pick the wrong first hire in a new function and set that function back a year. You’ll do a re-org that was correct on the whiteboard and was miserable to live through.

None of this is failure. It’s what scaling looks like from the inside. The teams that make it through are not the teams that avoid these things — they’re the teams that have a healthy enough culture to call them out, correct them fast, and keep moving. Which, coming back to it, is the whole game.