VSM Foundation cheat sheet

Value Stream Mapping originated in Toyota Production System (TPS) as material and information flow mapping, formalized by Rother and Shook in 1999.

Lean VSM uses takt time, cycle time, lead time, and percent complete and accurate (%C&A) to expose waste.

DevOps adaptation replaces physical inventory with work items (features, stories, changes) and information flow with toolchain handoffs.

Key evolution milestones: Lean VSM (1990s) → Agile/Lean software value stream mapping (2000s) → DevOps value stream management (2010s) → continuous value stream intelligence (2020s).

Value Stream Management (VSM) extends mapping to ongoing measurement, governance, and optimization across the delivery lifecycle.

The Flow Framework (Kersten, 2018) introduced flow items, flow metrics, and value stream networks as a management model.

SAFe and DevOps literature popularized value stream identification and mapping as a prerequisite for organizing around value.

Modern VSM integrates with CI/CD, observability, and work management tools to automate data collection rather than relying on manual sticky notes.

A value stream is the sequence of activities required to design, produce, and deliver a product or service to a customer.

Operational value streams deliver direct customer value; development value streams create and maintain operational value streams.

Trigger: a value stream starts with a defined triggering event (e.g., customer request, feature idea, incident).

Value: each step must contribute to the customer-defined outcome; steps that do not are waste candidates.

Flow: work should move without batching, waiting, or handoffs; identify where flow breaks.

Identification approach: start from the customer and work backward, or start from the trigger and follow the work forward.

Value stream boundaries: define entry (trigger) and exit (customer outcome) explicitly to avoid scope creep.

Distinguish value stream from organizational silo: a value stream crosses functions; a silo does not.

Use value stream canvas: name, trigger, customer, steps, owners, metrics, and pain points.

Map current state first (as-is), then design future state (to-be) with explicit improvement hypotheses.

Include three layers: information flow (requests, approvals, tool handoffs), process flow (steps), and timeline (lead time vs. value-added time).

Calculate flow efficiency = value-added time / total lead time × 100%.

Calculate %C&A (percent complete and accurate) at each handoff to expose rework loops.

Use swimlanes for roles or systems to reveal handoffs and queues.

Mark queues and waits explicitly; they are the largest contributors to lead time.

Annotate each step with owner, tool, cycle time, and wait time.

Future-state map should specify kaizen bursts (improvement actions) with owners and dates.

Toolchain integration enables automated flow of work items and events across planning, code, build, test, deploy, and operate stages.

Key integration points: work item tracking ↔ source control ↔ CI/CD ↔ test management ↔ monitoring/observability.

Use webhooks, APIs, and event streams to connect tools; avoid manual status updates.

Traceability: link commits to work items, builds to commits, deployments to builds, and incidents to deployments.

Data model: normalize work item IDs, timestamps, and statuses across tools to enable cross-tool analytics.

Security: use service accounts, scoped tokens, and audit logs for integrations.

Toolchain mapping should identify handoffs, data gaps, and manual steps that break flow.

Integration maturity levels: manual → automated status sync → event-driven → closed-loop feedback.

Lead time = time from work item creation (or commitment) to customer delivery.

Cycle time = time from work start to completion for a single item.

Throughput = number of items completed per unit time.

WIP = work items in progress; Little's Law: average lead time = average WIP / average throughput.

Flow efficiency = value-added time / total lead time × 100%.

Flow distribution = percentage of work in each flow item type (feature, defect, risk, debt).

Flow load = number of items in progress; high flow load increases lead time.

Flow velocity = throughput trend over time; flow time = lead time trend.

Defect escape rate and change failure rate measure quality of flow.

Deployment frequency and mean time to restore (MTTR) are delivery and stability indicators.

Inspection cadence: daily standups for flow, weekly flow review, quarterly value stream review.

Use value stream metrics dashboard to visualize lead time, throughput, WIP, and flow efficiency trends.

Identify bottlenecks by finding steps with highest wait time or lowest %C&A.

Apply theory of constraints: improve the bottleneck first; non-bottleneck improvements do not increase throughput.

Use cumulative flow diagrams (CFD) to spot WIP accumulation and flow blockage.

Run value stream mapping workshops with cross-functional participants to validate data.

Inspect handoffs: each handoff adds wait time and potential for information loss.

Use root cause analysis (5 whys, fishbone) on top flow blockers.

Compare current state to future state to measure improvement impact.

Organize teams around value streams rather than functional silos to reduce handoffs.

Each value stream should have a clear owner (value stream owner) accountable for flow and outcomes.

Team topologies: stream-aligned teams, platform teams, enabling teams, and complicated-subsystem teams.

Stream-aligned team is the primary team type; it owns a value stream end-to-end.

Platform teams provide self-service internal services to reduce cognitive load on stream-aligned teams.

Enabling teams help stream-aligned teams acquire capabilities temporarily.

Complicated-subsystem teams handle components requiring deep specialist knowledge.

Value stream funding: allocate budget to value streams, not projects, to enable continuous flow.

Governance: define value stream KPIs and review cadence; avoid local optimization.

Evolution is continuous: map → measure → optimize → re-map.

Use improvement katas: define target condition, current condition, obstacles, and next experiments.

Prioritize improvements by impact on lead time and flow efficiency.

Remove queues and batching first; they are the largest lead time contributors.

Automate handoffs and status updates to reduce manual coordination waste.

Introduce WIP limits to reduce flow load and improve throughput.

Use feedback loops (retrospectives, incident reviews) to adjust the value stream.

Measure improvement with before/after metrics; avoid anecdotal evidence.

Scale improvements: pilot in one value stream, then spread patterns to others.