As promised, I’m taking a break from writing about AI.

And as I was writing the draft for this week, I realized it’s too long and too important for just one newsletter - so I’m splitting across 2 weeks.

Last week, a client booked time with me to talk specifically about one thing: the bonus structure for a new hire who is starting soon.

As great talent gets more expensive, hard to find, and harder to keep, the bonus and general incentive comp conversation is more important than ever.

For this client in particular, this is their first time hiring a lifecycle marketing manager.

When they came to me, everything else besides the bonus was already set:

The candidate had been asking about the bonus since receiving the initial offer, and the CEO said they’d "figure it out", but the hire started in a week and they hadn't figured it out.

Understandably, the CEO had spent weeks deciding which role to hire, writing the job description, working with the recruiter, and running the interviews.

By the time the client found the person and checked the box, it felt like the hire itself was the win.

But the thing that's supposed to drive the behavior the CEO is paying for - the bonus comp - received the least amount of attention.

Their instinct was the same one I hear from lots of founders: "we'll just tie it to contribution margin."

This is when I said “Woah. Woah. Woah. Wait a second.” (verbatim)

“That's easy to say. It's hard to do.”

Before I get into my thoughts on bonuses and incentive comp in general, indulge me a second on a contrarian take.

Maybe the answer is to not offer a big bonus at all.

A couple years ago when I was researching for Ecom CFO’s culture deck, I came across the infamous Netflix culture deck and they have a section on compensation.

Their approach is to pay people an above-market — even uncomfortably high — salary. Hire people who are intrinsically motivated. Then trust them to do great work, because doing great work is who they are - and bonuses won’t change that.

It doesn’t work for everyone, but I see it inside my own firm.

Most of our controllers aren’t incredibly motivated by year-end incentive comp. Whether or not their “bonus” is $3K or $10K at the end of the year doesn't change how they work this Tuesday.

They're generally wired to make things reconcile to the penny, and when something's wrong it bothers them and they go fix it. A carrot dangling twelve months from now doesn’t change this.

Plenty of people will read that and call it the dumbest thing they've ever heard.

“Show me the incentive, I'll show you the outcome”. Fair. This is a tradeoff, not a rule.

But what I see founders ignoring too much is the physical and emotional work of executing the bonus.

Every bonus you offer is something you have to set a target for, build a report around, check in on through the year, and defend when it’s time for the payout.

No bonus means none of that work exists. It's the easiest comp you'll ever administer.

Most founders reach for a complicated bonus by default.

They never ask whether a simpler version, or none at all, would get the same behavior without all the headaches.

Assuming you are going to pay a bonus, consider it as a spectrum: no bonus, a simple bonus, a complex bonus.

Every step towards a complex bonus buys you tighter alignment, in theory, but it also costs you two things which are almost never priced in: management and data.

No bonus. Nothing to manage, nothing to reconcile, nothing to dispute. If you still want to reward people, and you should, keep it off the metrics.

A couple thousand bucks at the end of the year to cover their family's Christmas presents. Or a team trip with spouses included. At Ecom CFO we do annual trips, and for a remote team it's made a bigger impact than any number on a pay stub.

A simple bonus. One clean metric NOT named “contribution margin”. Examples may include email open rate, inventory turnover, NPS, or even revenue.

Both sides always know whether it's on track or off track, so there's less to manage and less wiggle room for misinterpretation.

I understand how impractical it may sound to incentivize on a single metric, but I would challenge you with the question: “What if you could only pick one?”

A complex bonus. The further down the P&L you go, the harder the bonus is to administer. Every layer below revenue adds surface area — more to manage, more to get wrong, more to argue about.

This is where contribution margin lives, along with most profit-based incentives. It takes the most data, the most management, and the most expectation-setting of anything on the spectrum.

It sounds simple. Rewarding real profit instead of activity is easy to say. But once you map out what the person can actually control, and the reporting needed to prove they hit the number, it can turn into a nightmare to manage.

Contribution margin raises a question revenue never did: what can this person actually control? What about the tariff refund that lands three months late?

Maybe they own pricing but have no say over cost of goods. Every one of those is more surface area and more conflict waiting to happen — and it usually happens at payout time, when memories are fuzzy.

A different client of ours agreed to pay their channel manager 2.5% of contribution margin, capped at $50,000 a year.

It seemed clean. He ran Amazon — a walled garden, his own platform. We'd pull the numbers straight out of it and everyone would agree on them.

Then the year happened.

They started running Meta ads into Amazon. Then Meta handed them a pile of free credits for doing something. Nobody told the channel manager — not then, not for almost a year.

So at bonus time, the question nobody had a clean answer for was this: do those credits count toward his number? All of them? Half? There was no clean way to split it.

That's the trap with contribution margin. It's technically knowable, but practically unknowable. Sure, you could sit down before the year starts and if-then yourself to death on every edge case. Nobody does. And it's a waste of time.

So we hit year-end with little documentation and a number nobody trusted. He came back with four things he disagreed with — starting with how the bonus was even calculated.

I had to prove the math to him. I agreed with one of his four points, pulled it out, and his bonus went up. But he was still nowhere near what he had in his head.

All year, he'd told himself he was getting close to $50,000. He landed at $22,500.

With a clean metric, there's nothing to decide — the number is the number. A messy one turns into a management decision: ultimately how nice or not nice am I?

I sent the owners the full write-up with two numbers — full benefit of the doubt, or a stricter read of the agreement. They landed in the middle. And in the employee's head, anything short of the full benefit of the doubt leaves a bad taste.

The channel manager ended up quitting in part because of this disagreement.

Some of it was the money. Most of it was the feeling that the goalposts had moved, when really they'd just never been clearly set.

It's possible to run a more complex bonus when the role calls for it.

But instead of immediately reaching for contribution margin, think about whether a clean, simple number or a generous salary with employee perks would do the same job with none of the overhead.

If one of your roles genuinely needs the complex end, there's a right way and a wrong way to build it.

Next week, in Part 2, I'll walk through how to implement a complex bonus without it blowing up on you.

Until then,

— Sam

Talk to me

My inbox is open and I love hearing from you - respond to this email with your questions, feedback, things I screwed up, or just because.

Looking for more? Some of our most popular posts:

Did someone forward this to you? If you like it, you can subscribe here