Incentivising Extreme Talent: How to Reward People Who Already Earn Too Much

This headline caught my eye recently: “AI talent wars lead to superstar salaries for top tech staff".

How on earth do you incentivise a 35-year-old already earning $4mn and no shortage of headhunter interest?

This is particularly acute in the AI and advanced tech sectors, where the gap between average and exceptional is enormous and well understood. Everyone knows the best engineers are worth 10x or 100x. And that’s exactly how they want to be paid.

At a certain level, the traditional levers - more salary, a bigger bonus, extra stock options - just don’t cut it. You're dealing with people who already have what they need. Moreover, these individuals often aren’t just executing tasks, they’re creating products, businesses, markets, or even categories. Their impact is exponential, not linear. And that makes it hard to benchmark them using traditional compensation frameworks.

What Else Makes It Hard?

Three reasons:

1. They’re already rich. Or at least highly paid. The marginal value of more cash is low - especially if it’s taxed at 47% in the UK.
2. They’re restless. The elite tend to value optionality, autonomy, and the chance to do something intellectually exciting or meaningful.
3. They understand value. These individuals are savvy. They can model risk-adjusted outcomes better than your CFO, so don’t expect to wow them with vanilla LTIPs or restricted stock.

So how do you design for them in the ‘real’ world? Let's give it a go.

The Design Challenge

First, I think you need to recognise that these individuals are not motivated by salary bands or annual bonus percentages. They want upside, freedom, ownership, impact etc. Any reward structure probably needs to reflect those ambitions.

Second, I think you need to accept that there will be an inevitable tension between trying to align their vision of value creation with a corporate structure that prefers predictability.

For example:

• Defining what “success” looks like when there’s no product yet (you still need credible and measurable targets).
• Modelling value for something that hasn’t been monetised (gulp).
• Designing equity or synthetic instruments that don’t rely on traditional trigger events (but what about shareholder alignment?).

So What Could Work?

For private companies, most traditional incentive plans hinge on a liquidity event such as an IPO or trade sale. But many of these superstars aren’t building towards a corporate exit. They’re building products, models or platforms. That’s their value. 

Therefore, incentivising this level of talent isn’t just about upside, it’s about alignment and access. I think you're trying to build reward structures that:

• Create credible upside (not just “trust us” equity).
• Preserve autonomy and ownership of impact.
• Respect their intelligence (e.g. don’t patronise with “everyone gets the same deal”).

All of which, I think, means incentives tied to milestone achievement work better than exit-driven schemes. For example: 

• Dynamic equity: Adjusts based on actual contribution and growth.
• Synthetic equity: Pegged to product P&Ls.
• Roll-forward mechanisms: Deferred bonuses tied to downstream product monetisation.
• Internal liquidity: Create ways to crystallise value before a corporate event.

Final Thoughts

The challenge with superstar incentives isn’t generosity, it’s imagination. You need to reward in a way that feels fair, flexible, and aligned with how they view value. That invariably means designing outside the usual playbook.

At Burges Salmon, we help high-growth companies, funds, and platforms design bespoke incentive structures that go beyond the usual toolkit. Our focus: real alignment, real outcomes, and enough flexibility to keep even the most restless talent committed.