One page per view and one model, in your browser. The sliders, the range and the reading never leave it, so the board costs nothing to run. The model is a month-by-month chain from the budget to the loans, the interest they pay and the protocol's revenue, with history behind it so deals already in the pipeline land in the first months.
The chain, month by month
Money. The sales budget hires people first, whole or by halves, and pays for programmes with what is left; at the plan, about four in five dollars are people. Token incentives come from the treasury.
First meetings. Programmes bring meetings, each extra unit fewer than the one before; outbound and inbound add their own, and a better known brand brings more from every source. There are only so many counterparties worth a meeting, so the share of meetings that are new falls as the year fills up.
Teams. A share of first meetings qualifies. Each team works only as many qualified opportunities a month as it has people for; the rest wait and go cold. A new hire reaches full speed after the ramp.
Deals. A share of worked opportunities signs. After the cycle the deal goes live; partner launches, bank pilots and chain deployments also wait for an integration engineer. A shorter partner time to launch moves every partner deal along faster from month one.
Volume. A live deal ramps to its size. Institutional balances leave a little each month; chain deposits drop when the chain's incentives end.
The book. Deposits meet borrowers. Up to 90 percent of new deposits can be lent, plus the idle liquidity already in the markets; a share of deposits with no borrower finds one anyway because rates fall.
Interest and revenue. Borrowers pay interest on the loans, more for a rate fixed for the term. The protocol keeps its fee when the fee switch is on; it is off in the plan, so the board leads with the interest and shows payback at a 10 percent fee. Both are run rates at month twelve.
The range
Each uncertain number has a low end, the board's value and a high end. The board runs the model 300 times; each time it draws every number from a triangle between its ends, on a log scale so it is as likely to land below the board's value as above it. The range is where 8 runs in 10 land. The difference a move makes is drawn the same way, in each run, so it has its own range; the ranking of the next money uses 60 runs for each place. Where a team is full, the range is lopsided: a better number cannot add deals the team has no time for, a worse one takes them away.
Read the move
The reading is written by the page from the model: the moves that matter most, the numbers they rest on, the uncertain number that moves the result the most (each one set to its low and its high end in turn), and a test for the lever behind the biggest move. Nothing is sent anywhere.
Guard rails
Every number on this page is invented. The shape of the protocol is real; the protocol is not.
The board counts only what the team adds in the twelve months. The book the protocol already has, and how the market moves it, is not the team's doing.
Volume from a partner, a curator or an issuer is counted at the share the board believes the team caused, not at what the dashboard shows.
The page runs on its own: no server, no account, no cookie. In production the parameters would come from the budget sheet, the volumes from the chain and the meetings from the sales system.