# Find out whether a missed decision is a report problem or a routing problem Manual worksheet; invented situation, no automatic product import. ## Owned work - State competing explanations — Product lead — Visibility failure and decision-route ambiguity have distinct disconfirming observations. - Observe recognition before prompting — Research owner — For each authorized context, record locating the ask, identifying its owner and proposed response separately. - Choose a bounded intervention test — Decision owner — The findings support a placement test, an ownership-route test or the specific next missing observation; prevalence remains unknown. ## Workflow questions ### Why not ask whether people like a dashboard? A preference does not tell you whether the request was hidden or its deciding owner was unclear. Observe those two checks separately before choosing the remedy. ### Do two reading contexts prove the problem is common? No. This protocol helps distinguish explanations in bounded contexts; estimate prevalence with a separately appropriate research design. ### How does TeamBoost help organize this? Keep the investigation owned in a task, preserve the separate observations in its context and report the resulting intervention choice. The guide does not establish native research analytics or automatic user observation. ## Job to be done When a decision request is missed, I want to test whether readers cannot find it or do not know it belongs to them so we choose the right intervention. Two readers miss a pending choice. One locates the ask when pointed to it but cannot tell who should decide; the other knows the owner but cannot find the ask in the weekly note. A new dashboard may address only one of these explanations. ### Recognizable symptoms - A visible request still has no recognizable accepting owner. - A known decision owner cannot locate the ask in the report. - One dashboard proposal is expected to fix two different observed failures. ### Shortcut that fails Counting dashboard preferences cannot distinguish missing visibility from missing decision responsibility. Observe recognition and action separately. ### Before Two people missed the decision. Build a dashboard so both notice it. ### After — invented worked answer Discovery protocol: in two authorized reading contexts, ask each reader to locate the pending choice, identify who can decide, and explain the next response. Record these observations separately without prompting the answer first. Finding an ask without knowing its owner weakens a pure visibility explanation; knowing the owner but failing to find the ask weakens a pure routing explanation. Compare these signals before choosing a clearer report placement, an explicit decision route, or one further targeted observation. These invented observations suggest hypotheses; they do not establish their prevalence or validate a feature. ### Conditional next moves - If The reader finds the ask but cannot identify its owner: Test a clearer decision route. Finding the request weakens a pure placement explanation. - If The reader knows the owner but cannot find the ask: Test clearer report placement. Known responsibility weakens a pure owner-recognition explanation. - If Both checks fail or prompting changes the answer: Collect one targeted unprompted observation. Do not choose a universal remedy from an ambiguous example. ### TeamBoost route — confirm account availability - Track the investigation question: Track the findability-versus-routing investigation and its intervention choice. Your check: The owner separates explanatory evidence from a feature commitment. - Preserve observations and hypotheses: Record find-the-ask, identify-the-owner and next-response observations separately for each authorized reading context. Your check: Preserve what happened before prompting; invented cases do not prove prevalence. - Report the resulting choice: Report which explanation weakened, the bounded test chosen and what remains unknown. Your check: Implementation and research permissions are agreed separately. ### Success check The investigation can distinguish finding the request from recognizing decision authority and choose a bounded intervention using evidence that could disconfirm either explanation. ### Editable draft facts - Intervention choice [choice]: Clearer report placement or explicit decision route — Name the alternatives that need evidence. - Competing explanations [explain]: Ask is hard to find; decision owner is hard to recognize — Separate explanations that require different remedies. - Unprompted observation steps [observe]: Locate ask, name deciding owner, explain next response in two authorized contexts — Record behavior before showing the answer. - Evidence against each explanation [disconfirm]: Finds ask but lacks owner weakens placement-only; knows owner but cannot find ask weakens routing-only — Say what would change your interpretation. - Bounded return decision [exit]: Choose one placement or route test, or the specific missing observation; no prevalence claim — Give research a decision and an uncertainty boundary. ### Draft pattern — replace every placeholder ```text Discovery observation protocol Intervention choice: {choice}. Competing explanations: {explain}. Before prompting, observe: {observe}. Disconfirming evidence: {disconfirm}. Return decision and limits: {exit}. Do not treat two illustrative contexts as validated demand or population prevalence. ``` ### Limits The worked case is illustrative, not measured customer demand. Product tracking does not replace a responsible research process. Local text preparation only; no workspace read, task creation, send or approval. Source interfaces do not establish account availability or live feature completion.