Facilitator Guide · Block 3.4 of 3.5 · Second Divergence

Refine the Roadmap and Test How the Value Stream Should Collaborate

Use Day 2 to close the most important roadmap gaps and frame changes to Solution Areas, ARTs, teams, platforms, communities, decision paths, and syncs as testable hypotheses. Involve the people affected, define what better Flow would look like, and route validation into the refinement cadence.

Fast exercise
5 minutes
Full exercise
25–45 minutes
Input
Adjusted draft + Day 2 focus
Output
Roadmap + operating-model hypotheses
Deck basis
Slides 57–58

Understanding Day 2 Breakouts

Roadmap choices and collaboration design shape each other

Timing, sequence, capacity, dependencies, ownership, and architecture often reveal that the current collaboration structure cannot support the intended Flow. Day 2 makes those tensions explicit without treating a workshop sketch as a finished reorganization.

Roadmap and Operating Model

Refine what must happen and how people must interact

Roadmap timingWhen does value or learning need to arrive?

Milestones, windows, sequence, horizon, integration points, and validation dates.

Work ownershipWho can carry the outcome end to end?

Teams, Solution Areas, ARTs, Solution Trains, platforms, partners, and named initiative owners.

DependenciesWhich interactions create Flow or waiting?

Handoffs, shared services, expertise, environments, evidence, supplier, and decision dependencies.

Decision systemWhere should choices be made?

Local authority, cross-unit agreement, centralized guardrails, escalation, and artifact ownership.

Interaction patternHow should units collaborate?

Temporary collaboration, service consumption, facilitation, community, integration, or sync.

Refinement pathHow does the hypothesis become real?

Affected-person validation, experiments, owners, measures, decision dates, and review cadence.

Roadmap gapFlow interaction problemOrganization hypothesisAffected-person validationRefinement experiment

Organize around Value as Hypothesis

Design for better Flow, then test the consequences

Organizing around value is not synonymous with drawing a new org chart. Start from a specific Flow problem, describe the proposed boundary or interaction change, state the expected outcome, and define how the people affected will challenge and validate it.

Flow problem

Which waiting, handoff, overload, queue, dependency, feedback, or decision delay should improve?

Proposed change

Boundary, team, Solution Area, ART, platform, community, role, interaction, or sync hypothesis.

Expected outcome

Faster decision, fewer handoffs, clearer ownership, lower cognitive load, better quality, or shorter feedback.

Affected people

Who performs, consumes, owns, supports, governs, supplies, or is constrained by the change?

Trade-offs and risk

New dependencies, capability gaps, local optimization, transition cost, ambiguity, or loss of resilience.

Validation

What evidence, simulation, conversation, experiment, measure, and review date would change confidence?

Organization hypothesis We believe changing [boundary or interaction] for [affected value flow] will improve [observable outcome], because [evidence]. We will test it with [affected people] through [experiment or refinement] by [date], while watching [risks].
CollaborationDiscover together for a bounded period

Use when several units need intensive joint learning or design; define the outcome and when the collaboration can end.

Service interactionConsume a clear capability

Use when a platform or shared service can reduce coordination through reliable ownership, interfaces, and experience.

FacilitatingHelp another unit become capable

Use when expertise should unblock or grow capability temporarily rather than become a permanent dependency.

Optional lens: These interaction modes can sharpen a hypothesis; they are not a requirement to adopt Team Topologies as a full operating model.

Running the Exercise

Five moves from adjusted draft to refinable collaboration hypotheses

  1. 1

    Approximately 4–8 minutes

    Select Day 2 Themes from the Planning Adjustment

    Choose the few gaps and hypotheses that most affect the draft roadmap. Carry forward unresolved Day 1 work only when the new guardrails or evidence make a materially better outcome possible.

    Roadmap Refinement

    Clarify timing, sequence, owners, milestones, confidence, and unresolved dependencies.

    Organize around Value

    Which collaboration structure would improve the targeted value Flow?

    Solution Area Adjustment

    Move, split, combine, or strengthen Solution Areas as a hypothesis.

    ART / Solution Train Cut

    Test whether trains still match the value stream, dependencies, and decision demand.

    Team Structure Hypothesis

    Should teams change, collaborate differently, or form temporary cross-unit groups?

    CoP / Community Design

    Architecture, DevOps, cyber, AI, compliance, platform, quality, or supplier communities.

    Platform Consumption

    How can units use common capabilities without blocking one another?

    Architecture / API Governance

    Who decides interfaces, standards, evolution paths, exceptions, and evidence?

    Refinement Cadence

    Which refinements happen before Final Sync and PI Planning?

    Sync Meeting Design

    Which syncs are decision and artifact-update events rather than status meetings?

    Operating Model Hypothesis

    How should the Value Stream collaborate during the next learning horizon?

    Validation Card

    Who affected by the hypothesis must inspect, challenge, and adapt it?

    Day 2 Theme Selection
    ThemeRoadmap gapFlow problemHypothesis neededDecision / validation windowExpected output
    Integration structureunclear sequence ownercross-ART waitingtemporary Solution AreaDay 2 14:00roadmap + experiment
    Platform usageadoption milestonesshared expert queueservice + enabling pathFinal Syncconsumption model
     
    Speaker Notes
    Facilitator intent

    Select Day 2 breakout themes that refine the roadmap and test collaboration structure based on Day 1 learning, not based on the original agenda alone.

    How to facilitate

    • Start from the reset Day 2 focus backlog and choose themes that close roadmap gaps, organization hypotheses, CoP needs, platform usage, decision model, or sync cadence.
    • For each theme, state the roadmap question, affected people, validation needed, and intended output.
    • Limit themes to what can materially improve the draft roadmap or operating model before Final Alignment.

    Listen and watch for

    • Day 2 themes chosen because they are popular, not because Day 1 changed the decision basis.
    • Roadmap refinement and organization design mixed without knowing which output is expected.
    • Too many themes, causing thin participation and weak artifact updates.

    Avoid

    • Treating Day 2 as a second Day 1. Day 2 should sharpen, validate, and make the draft usable.
    • Letting organizational design conversations run without a roadmap problem to solve.
    • Ignoring unresolved Day 1 evidence because Day 2 already had planned topics.
  2. 2

    Approximately 5–10 minutes

    Frame Organization Hypotheses from Flow Problems

    Prevent preferred structures from appearing as solutions without a problem. For every proposal, make the current Flow mechanism, expected improvement, affected people, trade-offs, and validation path visible.

    Problem-groundedNames an observable Flow mechanism

    Waiting, handoff, overload, queue, dependency, feedback, ownership, or decision delay.

    Outcome-orientedExplains what becomes better

    Decision speed, Flow time, quality, learning, capacity, customer feedback, or reliability.

    BoundedChanges the smallest useful unit

    Interaction, responsibility, temporary group, Solution Area, service, sync, or train boundary.

    Human-awareNames the people and work affected

    Include cognitive load, skills, identity, authority, career, support, and transition reality.

    Trade-off-awareShows what could worsen

    New coordination, capability gaps, local optimization, fragility, governance, and transition cost.

    TestableHas evidence and a review point

    Prototype, simulation, trial horizon, measure, affected-person feedback, and decision date.

    Organization Hypothesis Backlog
    Flow problemProposed changeExpected outcomeAffected peopleRiskTest / evidenceOwner + review
    integration decisions wait 12 daystemporary Integration Solution Areadecision within 3 daysART A/B, architectsextra meeting loadone-PI trialSTE · Final Sync
    platform team becomes ticket queueservice + enabling interactionfewer escalationsplatform + consumersconsumer skill gaptwo adoption slicesPlatform PM · 6 weeks
     
    Speaker Notes
    Facilitator intent

    Frame organize-around-value ideas as hypotheses to be tested with affected people, not as management decrees hidden inside a roadmap workshop.

    How to facilitate

    • Write the hypothesis as: if we change collaboration structure in this way, Flow improves for this objective or problem, because this interaction changes.
    • Name affected ARTs, Solution Areas, teams, CoPs, suppliers, platforms, and decision interfaces.
    • Define the validation needed: who must inspect the hypothesis, what evidence would support it, and what risk would make it unsafe.

    Listen and watch for

    • Structure changes proposed as obvious fixes without linking them to Flow problem, objective, or roadmap gap.
    • People affected by the change not present or not scheduled to validate it.
    • Permanent operating-model language used for what should be a PI-level experiment.

    Avoid

    • Finalizing organization design in the agenda design exercise.
    • Using the conference to impose a train cut, team structure, or CoP model without inspect-and-adapt path.
    • Treating every collaboration issue as an org-chart issue. Many are policy, cadence, artifact, or decision-right issues.
  3. 3

    Approximately 5–10 minutes

    Map the Interaction, Decision, Artifact, and Cadence Changes

    A structural label alone does not explain how work will move. Show who interacts, for what purpose, through which decisions and artifacts, with what duration, and how the interaction ends or evolves.

    Units

    Teams, Solution Areas, ARTs, trains, platforms, functions, partners, customers, and communities.

    Purpose

    Joint discovery, integration, service, enablement, governance, evidence, decision, or learning.

    Mode

    Collaboration, service, facilitation, community, decision window, standby, or direct handoff.

    Decision rights

    Local authority, shared agreement, centralized guardrails, escalation, and delegate.

    Artifacts

    Roadmap, backlog, interface, service contract, risk, decision log, board, or evidence chain.

    Cadence / trigger

    Continuous, weekly, per iteration, milestone, decision age, risk trigger, or temporary window.

    Success signal

    Waiting, rework, handoff, decision time, quality, adoption, feedback, or workload change.

    Evolution rule

    When should the interaction stop, change mode, expand, shrink, or become a stable capability?

    ART A + ART BTemporary collaborationIntegration decision logWeekly for 6 weeksReview handoff time
    Speaker Notes
    Facilitator intent

    Map the concrete interaction changes needed for roadmap refinement: which roles, artifacts, decisions, syncs, or handoffs must change for the hypothesis to work.

    How to facilitate

    • For each theme, name what changes in interaction: decision owner, artifact owner, sync cadence, interface rule, platform consumption, CoP involvement, or validation path.
    • Tie each interaction change to a roadmap entry or risk; otherwise it remains an operating-model wish.
    • Identify whether the change is a one-time conference output, a refinement cadence item, or an execution-cadence item.

    Listen and watch for

    • Generic collaboration language such as align better, stronger ownership, or more transparency without a changed interaction.
    • New syncs created without artifact update, decision right, or trigger. That becomes meeting load.
    • Interaction changes that depend on people or functions not in the validation path.

    Avoid

    • Over-designing the future operating model instead of mapping the next useful interaction change.
    • Creating a sync for every problem. Some problems need a decision rule, artifact owner, or clearer interface instead.
    • Ignoring the difference between refinement before PI Planning and execution realignment after PI Planning.
  4. 4

    Approximately 5–10 minutes

    Validate with the People Who Must Make the Hypothesis Work

    Invite affected people to inspect the Flow problem, proposed change, expected outcome, operational consequences, and experiment. Validation is not a ceremonial vote; it is evidence that can change the hypothesis.

    Work realityDoes the proposal match how value and information actually move?

    Surface hidden handoffs, support work, local constraints, and informal coordination.

    CapabilityCan the proposed unit perform the responsibility?

    Skills, capacity, tooling, decision authority, access, platform support, and learning needs.

    Cognitive loadDoes the change simplify or overload?

    Domain breadth, interruptions, coordination, operational ownership, and competing priorities.

    InterfacesWill boundaries become clearer?

    Inputs, outputs, quality, service expectations, escalation, decision rights, and ownership.

    TransitionWhat must change safely?

    Work in progress, responsibilities, careers, contracts, metrics, access, communication, and support.

    Learning designWhat evidence would change our mind?

    Experiment horizon, baseline, measure, review, stop rule, and supersession decision.

    Organization Hypothesis Validation Plan
    HypothesisPeople affectedValidation methodEvidence / measureOwnerBy whenDecision after validation
    Integration Solution AreaART A/B teams, architectsmapping + one-PI trialdecision age + handoffsSTEFinal VS Synccontinue / adapt / stop
    Platform service modelplatform + 3 consumer teamsservice prototypeadoption + support loadPlatform PM6 weeksscale / enable / revise
     
    Speaker Notes
    Facilitator intent

    Validate roadmap and operating-model hypotheses with the people affected before the conference treats them as credible draft decisions.

    How to facilitate

    • Bring affected people into the room or validation window: teams, ART reps, Solution Areas, CoP owners, suppliers, operations, and central functions where relevant.
    • Ask what would make the hypothesis wrong, unsafe, too expensive, or unworkable. Preserve dissent as evidence, not resistance.
    • Use confidence and validation status: accepted for draft, needs refinement, unsafe, or route to another decision body.

    Listen and watch for

    • Affected groups learning about the hypothesis only after Final Alignment.
    • Validation reduced to consent language, while feasibility, capacity, and quality concerns remain unresolved.
    • A high-confidence label based on leadership agreement rather than affected-person validation.

    Avoid

    • Equating stakeholder buy-in with validation. The question is whether the structure can work in the real work system.
    • Treating every objection as a blocker. Some objections are validation data for refinement.
    • Moving unvalidated organizational hypotheses into the roadmap as if they were confirmed decisions.
  5. 5

    Approximately 4–8 minutes

    Update the Roadmap and Operating-Model Hypothesis Together

    Capture the refined roadmap entry, collaboration hypothesis, owners, validation work, decision rights, cadence needs, risks, and review point in connected artifacts. The roadmap says what must move; the operating-model hypothesis says how the system will learn to move it.

    Roadmap deltaTiming, sequence, scope, owner

    Include assumptions, dependencies, capacity, confidence, and unresolved questions.

    Organization hypothesisBoundary or interaction change

    Connect the Flow problem, expected outcome, affected people, trade-offs, and test.

    Decision modelAuthority and escalation

    Clarify local choices, shared agreements, guardrails, delegates, and decision windows.

    Interaction mapPurpose, mode, artifact, duration

    Show how teams, trains, platforms, functions, partners, and communities collaborate.

    Refinement backlogValidation and implementation work

    Owners, minimum evidence, due dates, measures, risks, and next decision.

    Cadence needSync or trigger

    Route the work into Final Sync, PI Planning, community, architecture, supplier, or other refinement loops.

    Day 2 breakout handoff We refined [roadmap entry] and propose [organization hypothesis]. [Affected people] will validate it through [work] by [date]; [owner] will update [artifact], and [next sync] will decide [continue, adapt, scale, or stop].
    Day 2 Breakout release rule READY

    roadmap gaps and collaboration changes are connected, affected people can influence the hypothesis, and validation has owners, evidence, and a decision date

    NOT READY

    the room has drawn a future org chart, assigned people without them, or created syncs with no decision or artifact purpose

    Speaker Notes
    Facilitator intent

    Update the draft roadmap and operating-model hypothesis together so timing, ownership, refinement cadence, and validation remain connected.

    How to facilitate

    • For every Day 2 output, update the roadmap entry, operating-model note, owner, decision log, risk, and next refinement step.
    • Distinguish confirmed draft decisions from hypotheses that require validation after the conference.
    • Prepare the Final Alignment handoff: what changed, who owns it, confidence, risk, and what must be reviewed before PI Planning.

    Listen and watch for

    • Roadmap entries updated without the collaboration structure needed to make them real.
    • Operating-model decisions documented separately from the roadmap, causing them to be forgotten during refinement.
    • Outputs that lack owner, review date, or artifact path.

    Avoid

    • Claiming the roadmap is final. The deck frames the output as a draft for refinement.
    • Leaving structural hypotheses as whiteboard notes instead of artifact updates.
    • Entering Final Alignment with multiple conflicting versions of the draft.

Facilitation

Keep organizational design close to Flow, people, and evidence

IF THE ROOM STARTS WITH BOXES

Return to the Flow problem, value boundary, decisions, handoffs, waiting, cognitive load, and expected outcome.

IF A STRUCTURE IS TREATED AS FINAL

Rewrite it as a hypothesis with affected people, test horizon, success signal, review date, and stop rule.

IF PEOPLE ARE “MOVED” ON A CANVAS

Pause. Name who must validate, what work and identity are affected, and how the transition decision will be made.

IF EVERY PROBLEM CREATES A NEW SYNC

State the decision or artifact-update job, cadence trigger, participants, owner, and condition for merging or stopping it.

IF PLATFORM MEANS TICKET QUEUE

Inspect consumer needs, ownership, service experience, enablement, interfaces, and the interaction mode—not only team labels.

IF VALIDATION MEANS “DO YOU AGREE?”

Ask what evidence, risk, work reality, or consequence would change the hypothesis and who can provide it.

Ready to Move On?

Day 2 Breakouts are ready when roadmap and collaboration can be refined together

  • Day 2 themes come from the adjusted roadmap, unresolved decisions, and highest-leverage Flow problems.
  • Every organization proposal names the observable problem it is intended to improve.
  • Roadmap timing, ownership, dependencies, decisions, and refinement path are explicit.
  • Solution Area, ART, train, team, platform, community, and sync changes remain hypotheses.
  • Expected outcomes and possible negative trade-offs are visible.
  • Interaction purpose, mode, decision rights, artifacts, cadence, and evolution rule are defined.
  • The people affected by each hypothesis are identified and able to challenge it.
  • Capability, cognitive load, interfaces, transition, and learning design are considered.
  • Validation has a method, evidence, owner, date, and next decision.
  • Roadmap and operating-model artifacts are updated together.
  • Refinement work and cadence needs are routed to explicit next syncs.

Source Foundation

This page distills slides 57–58 of 05 Value_Stream_Conference.pptx. Organization Hypothesis Cards, interaction-mode lenses, affected-person validation, evolution rules, and connected roadmap/operating-model updates are facilitator extensions. Team Topologies is used selectively as an interaction-design lens rather than imposed as a second framework.