Two engineers, three roadmaps
Six people, seed round closed, revenue climbing.
Their fastest-growing acquisition path wasn't sales. It was a handful of fractional CFO firms who kept putting clients onto the platform. Nine of them by the spring, doing it entirely by hand: email a spreadsheet in, wait for someone to provision accounts, chase for the referral fee at quarter end.
The firms wanted a portal. The startup wanted to give them one. Their CTO ran the numbers on building it in-house and got an answer she hated: eleven weeks of one engineer, which meant the core product roadmap stopped dead for a quarter.
Hiring wasn't faster. A mid-level engineer in Austin was a six-week search minimum, then onboarding, then a $155k salary against a runway that had eighteen months on it.
The bit that convinced them
Six-day spec sprint, $3,200. The CTO expected a document. What she got back included a clickable prototype of the whole portal, which she forwarded to three of the partner firms before the build had started.
Two of them wrote back with changes. One wanted white-labelled client invitations. Another pointed out that partners at those firms rotate clients between them constantly, so a flat "partner owns account" permission model would break in week one.
That feedback arrived before a single line was written. Her estimate is that catching the permissions issue at prototype stage instead of at launch saved somewhere north of three weeks of rework.
The build, including the part that went sideways
Fixed price, $27,500, seven milestones, five weeks.
In week three they changed their mind about the permissions model again, moving from role-based to a delegated-access design. That's a scope change, and we priced it as one: $2,100 and four extra days, quoted and agreed in under 24 hours. No renegotiation, no "we'll see where we land."
The CTO's view: "Every agency I've worked with treats a change request as a negotiation. Getting a number back the next morning was genuinely disorienting."
Final: $29,600 and five and a half