AI evaluation's expert track commands almost seven times the no-experience entry rate

The climb contains two separate gates, and they are priced for different reasons. Moving from the $15 no-experience listing ^[3] to the $35 he reached after clearing the platform's assessments ^[5] is a factor of roughly 2.3 ^[3], and it is a test that a literate applicant can sit on a weekend. Moving from $35 to the $100 he has billed for ten months ^[9] is a factor of roughly 2.9 ^[2], and it rests on fifteen years of sales management ^[2] that no assessment can grant. The full trip is about 6.7 times the entry rate ^[1], but only the upper half is scarce, and that upper half is what a buyer is actually paying for.

The gate on that upper half is platform history rather than credentials. Zohourian says that once sales projects appeared, what put him on them was an accumulated record of high-quality tasks across the platforms he had already worked on, and that without that history he does not believe he would have been selected ^[8]. The expert tier, in other words, is filled by re-tiering people already inside the funnel. A buyer opening evaluation in an uncovered vertical is waiting on two clocks: how long it takes practitioners in that field to find the work at all, and how long it takes them to build enough task history to be trusted with the high-stakes version of it.

This is one contractor's essay, not a rate card. Business Insider states that it confirmed his work and project rates ^[1], and the parts that matter for procurement were designed by the buyers rather than by him. The platforms set the assessments and gate the tiers ^[5], and they price each project on its own terms, with no baseline and with some projects paying less than the one before ^[6]. What the record does not give us is how many workers clear the expert gate, what share of an expert's hours bill at the top band, or whether $100 holds. Those are longer-run questions; the hiring question is immediate, and it belongs to this quarter.

Evaluation costs are rising, and negotiating volume terms only goes so far, because the unit of work moved underneath the price: the early task was writing a prompt, two responses and a short justification, while the current agentic work is described as completely different and more complex ^[10]. Volume terms assume interchangeable units of labeling, and what is being bought here is a practitioner's read on whether a system behaved correctly in work they have personally done.

There is a second reason the labor line will not behave like a BPO contract. Zohourian's evaluation income is what allows his own ventures to keep their profits: he says the businesses are profitable, that he draws no salary from them, and that he lives on the evaluation work while reinvesting the rest ^[14]. At $100 an hour across 20 to 40 hours a week ^[9][11], that is $2,000 to $4,000 of weekly billing ^[4] underwriting a consultancy and a training platform ^[13]. He also reports pauses of a week to a month between projects ^[11], and a four-week gap forgoes $8,000 to $16,000 ^[5], which is the exact exposure that pushes a specialist toward income they control. The people best qualified to judge a sales agent are people with a firm of their own to build. Right now that looks like cheap access to expertise you could not otherwise hire, but it can turn into a renewal conversation with someone whose fallback is their own company.