Distorted Reality | Grandimam

← Grandimam

Suppose there are two engineers. One (EA) makes $200K; the other (EB) makes $50K. EA was a strong performer and EB solidly met expectations.

Now the CEO wants to update the company’s performance metrics, so he adds salary as a factor in how engineers are evaluated.

What do you think happens?

EA, the stronger engineer, is scoring below EB. When salary becomes the denominator, being expensive looks like underperforming, even though neither engineer changed the way they work.

On top of that, EB now believes he is a top performer, not realizing he is no longer being measured against the right standard. He is simply being rewarded for being cheap. EA is being told she is slipping when she is not. Both are misled, in opposite directions.

Performance is not value per dollar

There is a difference between how good an engineer is and how much value per dollar an engineer delivers. The first measures individual performance; the second measures a business call. Comp-per-dollar is a fine tool for budgeting, but an improper way to measure how good an engineer actually is.

The shift LLMs accelerated

This is the shift LLMs have accelerated. The metric quietly moved from impact to intelligence per dollar.

I am starting to suspect engineers might begin pricing themselves down just to meet or exceed expectations. But a metric that rewards being underpaid is not measuring performance. It is deriving a business value that engineers have very little control over. It is also why layoffs should not be taken personally, because the metric was never measuring value or impact in the first place.

The distortion

And that is the distortion. EA thinks she is slipping. EB thinks he is excelling. The org thinks its numbers are improving. Three different people, all reading the same broken metric, all convinced of something that is not true.