Harbor Workspace / fictional product01 / Direction

Planning cycle 04 · planning view

Less searching. More doing.

Help small project teams find a trustworthy answer, understand its owner and take the next useful action without switching tools.

Horizon and confidence indicate direction, not delivery commitments.

Harbor Workspace / fictional product02 / Outcomes

Start with the change we seek.

O1 / intended outcome

Find the current project decision

BASELINE92

TARGET60

median seconds per task

Synthetic baseline from 12 moderated tasks. Target is a planning hypothesis, not an achieved result.

O2 / intended outcome

Identify an accountable owner

BASELINE58%

TARGET80%

successful task completions

Synthetic task results; the target guides a future validation study.

Harbor Workspace / fictional product03 / User needs

Whose problem earns the effort?

New project contributor

Find the latest decision without reading a month of messages.

Fictional interview synthesis: decisions are scattered across notes, chat and issue comments.

Project coordinator

Know which decision needs an owner or a review.

Fictional workflow observation: stale status creates repeated clarification requests.

Part-time collaborator

Resume useful work after several days away.

Fictional diary study: contributors repeatedly reconstruct context before acting.

Harbor Workspace / fictional product04 / Prioritization principles

The rules behind the sequence.

01

Resolve a whole user question

Prioritize an end-to-end path from finding a decision to knowing what to do. Avoid isolated interface polish with no validated task benefit.

02

Prefer evidence over feature breadth

A smaller validated workflow is more useful than many untested controls. Attach each item to one observable outcome.

03

Make ownership visible

Information is only actionable when a contributor can identify who maintains it and when it was reviewed.

Harbor Workspace / fictional product05 / Now

01

Active focus

Now.

N1High confidence

Decision register

A searchable list of current decisions, their context and the person responsible for review.

Progression gate

Eight pilot users find the correct current decision in the moderated task.

OWNER / Core workflow team
OUTCOME / O1

N2Medium confidence

Owner & review state

Show a named owner and a review-needed state on each decision record.

Progression gate

Pilot contributors identify the correct owner without asking in chat.

OWNER / Collaboration team
OUTCOME / O2

Harbor Workspace / fictional product06 / Next

02

Sequence after learning

Next.

X1Medium confidence

Contextual decision links

Link project tasks to their relevant decision record using a simple explicit association.

Progression gate

Observe fewer context-reconstruction steps in a new project task.

OWNER / Core workflow team
OUTCOME / O1

X2Medium confidence

Review reminders

Prompt owners to confirm whether a decision remains current, with a quiet digest option.

Progression gate

Owners understand and complete the review without duplicate notifications.

OWNER / Collaboration team
OUTCOME / O2

Harbor Workspace / fictional product07 / Later

03

Exploration, not a promise

Later.

L1Low confidence

Cross-project reuse

Explore how a decision might be referenced by more than one project without losing its context.

Progression gate

Demonstrate a recurring cross-project need before committing to implementation.

OWNER / Discovery team
OUTCOME / O1

L2Low confidence

Suggested ownership

Investigate suggestions for missing owners while preserving explicit human confirmation.

Progression gate

Show useful suggestions with understood failure modes and an easy refusal path.

OWNER / Discovery team
OUTCOME / O2

Harbor Workspace / fictional product08 / Capacity assumptions

Leave room for the unexpected.

Core workflow team · working days

AVAILABLE40

/

PLANNED28

RESERVE / 12 DAYS

Twelve team-days remain for support, accessibility fixes and uncertainty discovered during the pilot.

Collaboration team · working days

AVAILABLE30

/

PLANNED21

RESERVE / 9 DAYS

Nine team-days remain for operational work and learning from the reminder experiment.

Team-day estimates are planning inputs. They do not imply a promised calendar release date.

Harbor Workspace / fictional product09 / Dependencies & uncertainty

Sequence is part of the strategy.

Item	Depends on	Horizon
N1	None in this map	Now
N2	None in this map	Now
X1	N1	Next
X2	N2	Next
L1	X1	Later
L2	N2	Later

Risks to revisit

A decision register can become stale if review ownership is unclear.
Reminder frequency may create noise instead of confidence.
Cross-project reuse may require stronger context and permissions than the pilot reveals.

Confidence is a supplied qualitative judgment. Dependencies express sequence, not elapsed time.

Harbor Workspace / fictional product10 / Learning experiments

Spend a little to learn a lot.

Experiment 01

A single current-decision view reduces search time.

Run the same realistic task with a prototype and a control workflow; record task completion and observed search steps.

Decision rule

Continue if the pilot shows a coherent task benefit without reducing answer accuracy.

Experiment 02

A quiet digest is sufficient for timely owner review.

Trial a weekly digest with a small consenting pilot group. Observe review completion and interview participants about noise.

Decision rule

Adjust or stop if reminders are routinely ignored or interpreted as duplicate obligations.

Harbor Workspace / fictional product11 / Explicitly deferred

A roadmap also says “not yet”.

Deferred

Automatic decision extraction

The current evidence does not justify the trust and correction burden of automatically publishing inferred decisions.

REVISIT / When the manual register reveals a repeated, measurable capture bottleneck.

Deferred

Custom dashboard builder

A flexible dashboard does not yet address the main search and ownership tasks.

REVISIT / When multiple validated workflows require different durable views.

Deferred

External marketplace

Integrations would widen support obligations before the core information model is proven.

REVISIT / After the core pilot establishes stable ownership and review semantics.

Harbor Workspace / fictional product12 / Decision gates

Keep the roadmap a conversation.

01

Approve a bounded pilot of the current-decision workflow.

OWNER / Product sponsor

Review observed task results after the pilot.

02

Keep Later items in discovery until their gates are satisfied.

OWNER / Product & engineering leads

Revisit at the next planning conversation.

Review the evidence at each gate. Move, narrow or stop work when the assumptions change.