Paying for Outliers: How the best startups think about compensation and employee value to secure outlier talent.

The idea that salaries or employee compensation is a 'special' kind of spend - one that has to be "fair," internally consistent, held inside tight bands, and moved on a 6 or 12 month cycle - is complete fiction.

This belief isn't reality - it is a workaround that enormous companies are forced into because, somewhere past a thousand-odd people, they lose the ability to see or accurately measure what any single person actually brings to the company.

Unable to judge individuals effectively, they reach for blanket guidelines and rules instead, and every one of those rules is a prosthetic for a company that has gone blind.

A startup isn't blind

In a startup, it's easy to see who carries the company and who gets carried. You can attribute impact to individuals, and you have all the information required to make pay and compensation choices based on that impact.

Instead, most startups decide to put on a blindfold, ignore impact entirely, and default to titles, years of experience, what the rest of the team is paid, how much a person got paid by their previous employer, when they last got a pay rise, and so on.

They use rules and guidelines to provide a solution to the problem they've never actually faced, and it is one of the most expensive mistakes that any startup will ever make.

Compensation rules are disadvantage

These rules and guidelines are only successful in preventing you from making decisions that help you win. An internal rule dictating that two people with the same title need to be paid the same only guarantees that you can never hire someone better than the people already working for you.

Yearly or twice yearly pay reviews ensure that you have a multi-month disadvantage in the fight to retain your best people, and nothing more.

Strict pay bands prevent you from latching onto hiring opportunities that can propel your company far beyond your competitors.

The very best people who deliver incredible amounts of impact are outliers. They do not fit neatly into the rules. Neither will their pay expectations.

It's literally just spend

There's absolutely nothing special about employee pay.

It's just spend, no different from any other. And normal spend does not have these arbitrary rules. It would be ridiculous if they did. (It has some rules, just not these rules).

Do you care if one marketing campaign cost more than the other when it has double the impact? If opening an office in a different location would bring in 5-10x the cost in value, would you really not do it? Have you never paid more for something better, because it'd cost you less than the cheaper option over the long run?

Employee compensation is exactly the same. It may feel different because of what it's paying for (humans) but it's not. And it needs to be thought of and assessed in the same way as any other measurable spend (because it is measurable spend), which is very simple to do with just three questions.

1. Can we afford this?
2. Will it move us forward in a way that justifies the cost?
3. Is there any better immediate option to spend this money on?

That's it. The other rules you have are not necessary, and it's the same whether you're investing in booth space at an event, or hiring a talented engineer. If the answer is yes to 1 and 2, and there's no immediate option that'd utilise that spend better - buy.

This has always been the case

Startups have thought wrong about comp for years, and it's not at all new, but the cost of getting it wrong was rarely fatal, just painful. But this cost has changed quite dramatically as of late.

In the past, a typical startup may fill their company with '2x' or '3x' people - meaning their typical employees have 2-3x the impact of the median employee across all companies everywhere, including corporates.

Some really great startup people might have 5x the median impact - or roughly double that of the average startup employee, and the truly exceptional would be the coveted "10x employee" that you've likely heard of, but may not have actually seen in the flesh due to their rarity.

Paying 20% or even 50% more for someone who does twice the job was always a good decision, provided you could make use of that additional impact. Netflix proved this quite famously - but most successful startups that you know also learned this quite early on.

But many startups would shy away from this, citing the disadvantageous rules that they have in place about parity or fairness as the reason.

In 2026, this is going to be a more costly mistake than ever before.

It's even more important now

With AI, there are people out there who can contribute well in excess of 10x the median employee's impact. Real examples of people who bring 20x or 40x the impact exist right now. These numbers might sound ridiculous, but they are not.

These people may cost 50%, 70% or even 100% more. We have seen examples of Sydney based startups doubling the salary of top performers, or going to market for new hires with salaries that would triple what some employees with the same title currently earn.

You may feel that this is reckless - but it's not.

Because spending 200% more for someone who removes the need to hire five more people is logical. This is smart deployment of capital, not reckless. In fact, choosing to hire five people when one could do the job for half the spend would be reckless - but it's what many startups are still choosing to do.

And truthfully, it would be naive to think you can compete with a team who spends the same or less overall, but that is an order of magnitude more impactful. But your compensation rules force you to.

It cuts both ways

None of this is arguing to simply "pay your people more". As mentioned, startups are not blind, and you should not make decisions around spend blindly. It's arguing that you should pay based on the impact the individual has for the company, regardless of what that might be, with a general aim to hire and employ people who have outlier levels of impact.

But this is a two-way street. If a person's impact hasn't grown, there's no reason their pay should. And if someone receives an offer from elsewhere that wouldn't make sense for you given the impact they provide, you wish them well and let them take it.

Currently, the most cost effective startup teams in true runway terms tend to consist of very well paid individuals who deliver massive impact, meaning these startups can achieve success with much lower headcount than average.

If you look at the trend in the headcount to revenue ratios at almost all succeeding startups currently, you'll see quite clearly what a successful startup team looks like in 2026.

But where one person may deserve double their salary in this new way of thinking - others may only justify half. A small team of outliers only stays small, and only stays a team of outliers, if you're willing to part with the people who don't belong in it.

Netflix made this famous with the keeper test - if someone on your team told you tomorrow they were leaving, would you fight to keep them? If the answer isn't an easy yes - well, not everyone's cut out for startups.

Dealing with pushback

Sometimes making the decision to pay based on impact isn't actually that hard - it's the judgement, pushback or ridicule from others that surrounds it. If you're hiring people who have mega impact (you should be), and you determine salary based on impact, these numbers are almost certainly going to look high to others.

Your board, an advisor, your VC, your head of people, other team members, whoever it might be - they may not like this approach. Bluntly, they should stop making your decisions for you unless they are prepared to take ownership of the cost their decisions create.

But, the reality is that they are used to dealing in averages. Benchmarks made up of market average salaries paid by market average startups. Perhaps your investors only have visibility of a market average portfolio that generates market average returns. If you want to replicate those average results by hiring average people and building an average startup, then you can absolutely think about compensation in a very market average way.

The average startup fails, by the way.

Either way, it's your call to make, not theirs.