Client wants my seat rate cut to $9 because "the AI does the dialing now"

At $9 with $7 loaded cost and $180 a month of tooling, you're at roughly $0.90 an hour of contribution before any supervision time, and supervision is real once AI output needs review. That's not a price, that's a favor with an invoice attached.

The claim you're being asked to accept is that the demo's performance transfers to this client's data. AI SDR demos run on clean, opted-in lists. Your client's list is skip-traced, and if 40 to 60% of the numbers are wrong or disconnected, the tool's advantage shrinks to the dialing itself, which was never the expensive part. Ask them to run the tool on 500 records of their own list and compare connected conversations per hundred against what your seat produces. Either the number supports their ask or it doesn't, and you stop arguing about philosophy.

Unbundling the tooling is the cleaner first move. Show the $180 as a line and price labor at what labor costs. It reframes the conversation from your margin to their stack, and it makes the second client's eventual comparison easier to explain because the rate card is the same for both.

Outcome pricing does something you may not want. Per appointment moves list quality risk onto you while payroll stays fixed, so a bad data month hits your payroll rather than their budget. If you go there, cap it: a floor per seat plus a bonus per qualified appointment, with "qualified" defined in writing before the first invoice.

Separately, eight months of market knowledge living in one person's head is a single point of failure regardless of how this pricing lands. Get the street-level notes into your own documentation while she's still on the seat.

I hold rate and offer scope instead. If a client wants to pay less, they get fewer hours, not a cheaper hour. Cutting the hour teaches them the number was soft, and the next conversation starts at $9 too. What I'd want in your position is the tooling out of the seat rate immediately, because you're currently eating a vendor bill that the client thinks is a feature of your labor.

Dumb question on the $180. Is that per seat or shared across seats? If it's shared, unbundling gets awkward when you have to explain why one client pays the whole dialer bill.

Per appointment pricing would be useless for my side of things. Nobody pays per maintenance ticket triaged correctly. Hours are the only unit that maps to tenant work, so if solder's agency moves everything to outcomes it stops being able to serve people like me.

Careful with per-appointment. Depending on the state and on who the VA is talking to, compensation tied to producing a specific transaction-adjacent result can start looking like paying an unlicensed person for brokerage activity, and the line varies state to state. Worth putting the exact structure in front of an attorney before it's in a contract, not after the first invoice.

As a client I'd take a flat monthly retainer over either. I want a predictable number, not a bill that moves with how many appointments happened.

The transition is what I'd worry about. Say you unbundle and the rate card now reads $13 labor plus $180 tooling. To the client that's the same money and an extra line item, so what stops them from saying fine, drop the tooling, we'll buy the dialer ourselves and you just staff it? Then you've lost the piece you wanted paid for and you're a pure labor vendor competing on hourly rate again. Not sure how you avoid that once the number is visible.