40 leads and zero contracts tells you very little. At normal conversion for cold-sourced seller leads you'd want somewhere north of 100 before the number means anything, and a run of 40 with a couple of near-misses on price is inside the range of ordinary variance. Deciding your pricing model on that sample is the first thing that breaks.
The second is that hourly moves the risk onto you completely. At $9 an hour, a caller doing 15 to 20 live owner conversations a shift, you're paying for output you can't verify without listening to recordings. Budget your own hours for that, or a QA line item, because the incentive under hourly is time spent rather than conversations had. The $200 signing bonus doesn't fix it, since a caller has almost no influence over whether you and a seller agree on price weeks later.
On the price objection: go back through the two that died and check what number the caller reported versus what the seller told you. If the caller logged a price expectation and it was already 15 percent above your max, the lead cleared a definition it shouldn't have, and that's a definition problem you can fix on either pricing model. Tighten it to a stated number plus a timeline plus condition, and add a rejection process where you kick leads back within 48 hours.
The piece your plan hasn't accounted for is that owning the script and the callers means owning the compliance exposure directly. Scrubbing, calling windows, and suppression become yours to run, and state calling statutes differ from the federal layer, so what your vendor was absorbing lands on your desk. Price that in before you compare $9 an hour to $28 a lead.
