Committed pricing is sold as a discount, and it genuinely is one. It is also a bet on your client retention, placed by you, and underwritten by you.
The vendor is not taking a position on whether you keep your book. They have priced as though you will, and structured the contract so that the consequence of being wrong sits on your side.
The offboarding case
Every MSP models the growth case before signing, because that is the case the discount is designed around. Far fewer model the other one.
This is not an argument against committed pricing
For a stable book with predictable growth, a multi-year commitment is often the right call and the saving is real. The argument is narrower: model the case where you lose your largest client, and know the number before you sign rather than afterwards.
It is also worth asking what happens on the upside. Some agreements ratchet — you can grow into a better rate but not shrink out of a worse one. That asymmetry is the part to read carefully, because it is where the risk transfer actually lives.
Ask the vendor to model the offboarding case with you. A vendor unwilling to run that scenario has told you which direction the contract is designed to work in.
What aggregate bands do differently
Aggregate seat bands price against the total book rather than a promised figure. Grow and the rate per seat falls; shrink and the bill shrinks with it. There is no seat minimum and no multi-year lock, and month-to-month is available.
The trade is that the headline rate at a given volume may be less aggressive than a committed equivalent. What you buy with that difference is the ability to lose a client without also losing money on the seats they took with them.