Updated 7 October 2026. The summary used to say Accord Book is cheaper than the status quo even before you count the subscription. That sentence is not a comparison. The $1,500 senior-time figure below is the one this page actually calculates. The $3,000 and the $20,000 are assumptions, and the meeting hours are not recovered in full by a portal.
Pricing pages list numbers. They rarely show you what you are already spending.
Before the prices, here is the one figure this page calculates, and the two figures that are only assumptions.
The onboarding estimate: $1,500 calculated, $3,000 assumed
When a new developer joins your team, the first four to six weeks are expensive in a specific way. They do not know the history. They ask senior engineers questions that were already answered in a Slack thread eight months ago — questions about why the auth service is structured the way it is, why you chose that database, why that one endpoint behaves differently.
The figure this page can calculate: a senior engineer earning $120,000 a year, spending 15% of the first month on those questions, is $1,500. 0.15 × (120000 / 12) = 1500. Anything past that, including productivity drag on the new hire, is an assumption. Calling the total $3,000 was that assumption. The card above states the calculated figure.
Accord Book gives new hires read access to the project. They can ask the project instead of interrupting a senior engineer for history that was already captured. That does not set the value of a hire at $3,000.
Status calls: about 200 hours, not $20,000 of waste
Four client status calls a week, an hour each, is about 200 hours a year (4 × 52 = 208). At a blended rate of $100 an hour, that time costs $20,000 if you price every hour that way. Those calls are how a client relationship works. They are not unbilled waste, and a portal does not delete them.
What the portal can change is the prep: someone compiling, by hand, a status the project record already has. A client with portal access can read project state without pulling a person out of the work to assemble it. That is a shorter meeting, or a meeting that starts from a shared record. It is not a $20,000 credit against the subscription.
Decisions made twice
Teams pay for decisions they already made when the rationale was never written down. The caching strategy was chosen in Q2. By Q4 the person who chose it may be gone, and the team spends another meeting on the same question.
Accord Book stores the decision with provenance: who said it, which thread or meeting it came from, and when. The next time the question comes up, the record is there to read. That does not mean the second meeting never happens. It means the meeting can start from the earlier rationale instead of from memory.
The decision fatigue page is not the evidence for this. The product claim is the lost rationale, not a general law about how decision quality decays through the day.
What Accord Book actually costs
Studio (small agency, up to ~8 people): $5,500/year + $1,500 one-time setup. Founding-cohort price: $3,300 Year 1.
Agency (mid-size): $13,500/year + $2,500 one-time setup. Founding-cohort price: $8,100 Year 1.
No per-seat charges. No usage metering. You bring your own API keys. The provider bills you. We do not mark that usage up.
The $1,500 figure is the calculated senior-time cost in the first month of one hire. The $3,000 and the $20,000 above it are assumptions. Do not add them and call the sum money the subscription returns.
The founding pilot
We are currently accepting three founding-cohort deployments. The pilot is six to eight weeks, assisted setup included, at no charge. Founding pricing locks in at signing and does not change.
Three spots. See what the pilot includes →