Keep implementation cost out of the identification stage
Aliases: cost as early filter · separate identification from delivery · feasibility later
What it is
Deferred cost in opportunity identification means: when research is being reduced to a list of opportunities, answer only who is underserved and on what outcome, and how large the gap is. Development effort, technical debt, and scheduling difficulty belong on the roadmap; they should not decide which problems are allowed to be seen. Filtering with “that would be expensive” too early turns the list into a map of cheap work rather than a map of problems.
Why it happens
Identification and delivery draw on different information. Identification rests on observation: who is stuck, on which kind of outcome, and whether the evidence holds. Cost rests on implementation knowledge: current architecture, staff skill, dependencies, compliance—often only oral estimates in a synthesis meeting. Mixed into one vote, the vaguest cost figure becomes the strongest veto, because “too expensive” cannot be rebutted by research material. Teams also mistake the cost of a half-formed solution for the cost of the problem, and sentence the problem before it has been named.
Studying it
Split the synthesis record into two passes. The first produces opportunities and evidence of user consequence only; person-months and technical feasibility are banned from notes and tables. The second invites engineering or operations to give order-of-magnitude estimates, each tagged with the solution form it assumes. Compare the two lists: which high-evidence items vanish once cost enters, and whether the vanishing rests on measured effort or on-the-spot intuition. If the research corpus contains no implementation information, mark every cost judgment as an unverified assumption.
Where it stops holding
Cost is forced in when the issue is clear illegality, a safety shutdown, or a regulatory deadline already committed—whether to act is no longer a research question. Tiny teams may not have a separate identification stage. Deferral is not a vow to ignore cost forever; it is a refusal to let unverified difficulty filter the field of view before the evidence is in. Treating every expensive issue as more real is equally a substitution of cost for evidence.
Applying it
- Lock the cost column on the opportunity worksheet during identification, or move it to a separate delivery-assessment sheet.
- When someone says “we can’t build that,” record the solution they assumed; do not delete the opportunity.
- Open roadmap talks with the cost-free ranking, then overlay estimates so deletions are visible.
- Audit items vetoed by cost: if the reason cites no effort figure or constraint source, return the item to the problem list.