Performance, Promotion and Pay: There's Always a Negotiation

At two different points in my career, at two different types of company, I had the same conversation.

The first was at a corporate. My manager sat down across from me with a folder in his hand. He showed me my band, explained how my rating had been arrived at in the calibration meeting, and then said the line: “The budget is closed this quarter.” He was polite, he was prepared, he wasn’t unkind. He spoke the language of the system fluently.

The second was at an owner-run company. I knocked on the door and raised the subject. He looked at me, smiled, and said one sentence: “You’re good, we’ll sort it out.”

Both produced the same result. In both cases the raise came in below what I’d expected.

Years later I realised the real lesson wasn’t about which company was more generous. Having a system didn’t change the outcome; it only changed the language of the negotiation. In the first post of this series I defined being corporate as “the decision can be separated from the person.” When money is on the table, that definition strips down to its barest form: on one side there’s a spreadsheet behind the decision, on the other there’s a memory.

Corporate: the system protects the average and punishes the edges

A corporate raise process has three parts: a band (the salary range for your role), a rating (your performance review), and a calibration. Calibration is the meeting where ratings handed out by different managers get pulled onto the same scale — meaning your rating takes its final form in a room full of people who have never met you.

I was wrong about the purpose of this system for years. I thought it produced fairness. It doesn’t.

What it produces is defensibility. Every decision needs a spreadsheet behind it so nobody can say “you played favourites,” nobody can sue, nobody can allege discrimination. That isn’t a bad goal; in a large organisation it’s genuinely necessary. But it has a side effect, and the side effect is precisely your problem:

Extraordinary work cannot be rewarded extraordinarily — because the spreadsheet doesn’t allow it. If there’s forced distribution, the number of people who can land in the top bracket is fixed in advance. If the band has a ceiling, nothing you do puts you above that ceiling. The system is built to protect the average; it files down both edges with equal care. It doesn’t fire the bad performer, and it doesn’t let the extraordinary one fly.

The conclusion to draw from this isn’t a complaint, it’s a tactic. The person who will defend you in the calibration meeting is speaking in a room you aren’t in. And he speaks with whatever he happens to be holding. Not with his memory — with his material.

Handing him that material is your job.

Your job isn’t only to do good work; it’s to translate good work into a language the system can read. “I stabilised that service” doesn’t make it onto the spreadsheet. How many people were affected, which decision avoided which cost, who works faster because of you — those do.

Owner-run: no system, so the story is everything

In an owner-run company there is no band, no calibration, no spreadsheet. A raise is the sum of three things: the memory of the last three months, the perception of loyalty, and the cash on hand that day.

The first time I noticed this, it made me angry. Because in that arrangement proximity is leverage: the person in the owner’s line of sight, the one who reminds him he put out a fire last quarter, is in a better position than the person who has quietly held the system up for three years. That’s unfair.

And calling it unfair doesn’t make it go away.

Once I understood that anger wasn’t getting me anywhere, I changed my approach. Nobody here is keeping a record of your contribution — so reminding them is your job. Recency bias, our habit of weighting the most recent thing most heavily, stops being an enemy and becomes an instrument: in a pay conversation, the three months before the conversation matter as much as the conversation itself.

This isn’t manipulation. Manipulation is making it look like you did work you didn’t do. Making sure the work you actually did is visible is simply what survival looks like in a system that keeps no records.

Why title inflation depends on the type of company

I’ve seen two business cards with “senior” on them at the two types of company, and they weren’t describing the same thing. The reason is simple: a title costs a different amount in each place.

	Owner-run	Corporate
Cost of a title	Zero	Tied to a band and a salary range
When it’s given	When money can’t be	When budget and the approval chain allow
What it tells you	How the owner feels about you	Which box the company has put you in
Is it easy	A one-sentence decision	A process with many signatures

In an owner-run company a title is free. You give a title to the person you can’t pay; in the short term it even works. In a corporate a title is expensive, because it’s tied to a band, a salary range and an approval chain — which is why it’s given reluctantly, and why it means something when it is.

The practical upshot: don’t be proud of your title without knowing what it says, and don’t be hurt by it without knowing what it doesn’t. I’ve written before that seniority doesn’t arrive with the title; the type of company only widens that gap.

The honest part: what nobody tells you

Now I’m going to write the most uncomfortable sentence in this post, and I’m writing it as an observation, not as advice:

At both types of company, the highest-return move is usually not an internal raise. It’s changing companies.

The reason isn’t bad faith, it’s structure. In a corporate, an internal raise is tied to a band and a budget cycle; an outside offer is tied to the market. In an owner-run company, an internal raise is tied to cash and to memory; an outside offer is, again, tied to the market. In both places your price on the inside is anchored to your price on the day they hired you.

But I’d be lying if I left that observation here without writing down its cost. When you change companies you lose things too, and none of them show up on a payslip:

• Context. You know why a piece of code was written the way it was, which fire a given decision came out of. That knowledge doesn’t travel; in the new place it accumulates from zero.
• Trust. “If he said he’d do it, he’ll do it” is built over years, not on day one.
• Standing. The weight your word carries in a meeting comes from your history in that room. In the new room, that history doesn’t exist.

These aren’t money, but they are a price. And whether that price is worth paying is a separate question — one I’ve discussed in the post on staying or leaving. What I’m saying here is narrower: don’t imagine you’re negotiating a raise without the possibility of leaving. The other side is already factoring that possibility in.

Question: Is it easier to get a raise at a corporate or at an owner-run company?

Answer: Neither. At a corporate a raise is a matter of proving; at an owner-run company it’s a matter of reminding. Having a system doesn’t make the raise easier; it only determines which language you have to speak.

Three desks

From the senior engineer’s desk. You have to put your contribution on the record. Trust the record, not the memory — because your memory is wrong and your manager’s memory is wrong, and they’re wrong in different directions. Your constraint: keeping records feels like self-promotion, and most of the good engineers I’ve known are repelled by it. I understand; I was repelled by it too. But I’ve also seen this: you’re the one who pays for that distaste. Nobody keeps the record on your behalf.

From the manager’s desk. On a fixed budget you’re playing a zero-sum game: what you give to one person, you take from another. You need data to defend your team and most of the time you don’t have it — because you can’t actually see what everyone is doing. Your constraint: in the calibration room, saying “she deserves it” isn’t enough, you’re asked to prove it. And the person who has to hand you that proof is the engineer herself. If she doesn’t, your good intentions won’t save her.

From the owner’s desk. “He deserves it” and “I can pay it” are two separate sentences, and they shouldn’t be confused. The most expensive lesson I learned once I sat at this desk was this: what breaks trust isn’t failing to give the raise, it’s not saying so. Because the sentence “you deserve it, but I can’t do it right now; I’ll do it on this date, if this condition is met” never gets said, people conclude they aren’t valued, and they leave. When in fact most of them leave not from feeling unvalued but from uncertainty. Your constraint: the cash is real, and a raise is a permanent commitment to an expense — once you’ve given it, you can’t take it back. That’s why you hesitate, and your hesitation looks like silence.

Closing

I didn’t write this post angry, because there’s no return on anger here.

Sulking at the system in a company that has one, expecting fairness in a company that doesn’t — these are two faces of the same mistake. I’ve sat at both desks and I’ve seen this: refusing to learn the language of the place you’re in is the quietest and most expensive waste in a career.

A raise isn’t a reward. It’s a negotiation. And the negotiation has already started, whether you’ve noticed or not.