Paying experts fairly is harder than it looks

Setting rates is the part of running an expert network where everyone has an opinion and almost nobody has the data. The temptation is to copy whatever the market is doing, shave a few percent, and hope nobody notices. We have done the opposite, deliberately, and this post is how.

Three forces set a Lona rate: market signal, project stakes, and tier multiplier. We publish all three in advance. Nothing in the payout formula is a secret.

Market signal

For each domain we collect a quarterly rate band from comparable freelance markets, agency placements, and direct contractor offers. We publish the band. Then we anchor the senior tier near the top of it — not the median, not the bottom. The reason is simple: cheap senior rates do not return senior talent.

Project stakes

A red-team task on a clinical decision-support model is not the same as a syntax-cleanup task on a code corpus. Both might take an hour. Only one carries real downstream consequence. Stakes-aware pricing means projects with higher consequence pay more, and the multiplier is on the project page before anyone claims work.

Tier multipliers

Quality compounds. An expert with two years of accepted submissions and a calibrated review record is, mechanically, a better bet for the next batch. The tier multiplier converts that track record into a higher rate on the same task. We publish the multiplier table; experts can see, exactly, what tier they are and what their next tier pays.

• Practitioner — entry tier, full published rate.
• Senior — 1.25× multiplier, requires sustained quality at Practitioner.
• Specialist — 1.5× multiplier, scarce-skill or regulated-domain certification.
• Reviewer — 1.75× multiplier, calibrated review record + senior arbitration role.

What we got wrong

Our first multiplier table was too compressed. The gap between Practitioner and Senior was 10%, which was not enough to motivate the work it takes to get there. The current 25% step is the result of looking at retention and tier progression numbers over six months and admitting we had cheaped out on our own incentive design. Saying it out loud is part of paying fairly too.