118 Must-Haves on a Fixed Budget
A membership association is building a new member portal with a fixed budget of $420k. The vendor's estimate for the 140 requirements gathered so far is $610k. Go-live is tied to the annual conference in seven months and cannot move. In the prioritisation workshop, 118 of the 140 requirements were marked 'Must'. The sponsor says nothing can be cut because 'our members asked for all of it', and points out that the requirements came from a member survey with 2,100 responses. Two prior portal projects were cancelled after scope disputes. The vendor needs a scope decision in ten days to hold the price.
1. How do you get from 118 must-haves to a viable scope without a confrontation?
Read the worked answer
118 Musts is not a prioritisation; it is a list. MoSCoW failed here because 'Must' was used to mean 'I want it' rather than 'the release fails without it'. Change the question.
Ask, for each requirement: if this is absent on day one, what specifically happens? Four consequence categories: a legal or regulatory breach; a member-facing process that cannot complete; a workaround with a named cost; or 'someone is disappointed'. Only the first two are candidates for a true Must. With the right six people in the room, 140 items can be triaged in about three hours.
Then:
- Group by theme - join and renew, profile and preferences, events, payments - so trade-offs are between coherent slices rather than 140 atomised votes.
- Put the money on the table. $420k buys roughly 96 of the 140 at the vendor's rate. Show the gap as a bar, not a sentence.
- Score silently and independently first, so the sponsor and the loudest director cannot anchor the room.
- Reframe from 'cut' to 'release 1 / release 2' with a dated roadmap. Nothing is removed; it is sequenced.
2. How do you make the resulting scope decision defensible to the sponsor and to the organisation?
Read the worked answer
A defensible decision needs four things: traceability to the source, an explicit criterion, a named decision maker, and a record.
- Link every requirement to its survey evidence. 2,100 responses does not mean 2,100 people said each thing. Attach response counts and segments - several 'musts' will trace back to a single open-text comment.
- Publish the criterion before applying it, so it cannot be accused of being retrofitted to a preferred outcome.
- Have the sponsor sign the release-1 scope and the sequenced roadmap, not a 'cut list'. The sponsor owns the trade-off; the analyst owns the evidence and the record.
- Keep a decision log: item, decision, rationale, decision owner, date, and dissent noted.
- Agree the revisit rule up front - what would move a release-2 item earlier (budget increase, date change).
Given two prior cancellations in this organisation's history, the audit trail is not bureaucracy. It is the mechanism that stops the same scope fight from restarting at release 2 - and it is what protects the sponsor as much as the project.