A technical founder does not need a board to feel validated; they need people who discuss hard decisions with context and no sales agenda. A badly built technical advisory board turns into four meetings a year of nice slides. With a contract, an agenda and a decision log, the same format becomes the place where the founder takes the architecture decisions that affect fundraising and product.
Three people with different roles
One on product, one on architecture, one on organizational scale. Ben Horowitz, in The Hard Thing About Things (HarperBusiness, 2014), reports that boards work when each member owns a domain and has the confidence to disagree. Four people with the same profile buy the same opinion four times, and the meeting turns into a round of agreement.
A 12-month contract with scope
How many meetings, which decisions, rate per hour. Short scope: one architecture review per quarter, two hours of availability for a hiring or vendor decision, a reply to a message within two business days. Without scope, the advisor becomes a contact you never use and the frustration runs both ways. Pay per meeting, with an annual cap, so the cost stays visible to you and to the board.
A fixed 90-minute agenda
Three blocks: pending decision, risk, learning. The decision block comes first and takes half the time, with two options and the founder's recommendation written beforehand. Marty Cagan, in Inspired (Wiley, 2017), insists that product decisions run on evidence; the board exists to check the quality of the evidence, not to pick the next feature.
A decision record that outlives the meeting
Five lines sent the same day. What was decided, what stayed open, who acts by when. Will Larson, in Staff Engineer (O'Reilly, 2021), describes writing decisions as the tool of influence for people who do not give orders; on a board, the record also keeps the same discussion from returning next quarter.
A criterion to renew or end it
Count the decisions you joined per year. If in 12 months the board took part in fewer than three relevant decisions, change the composition or end the contract. Renewing out of habit costs money and meetings the founder never gets back. Also record how many messages the advisor answered outside the meeting, and count that at renewal.
