Innovation Framework
Dimension 02Innovation Framework
Structure Innovation So It Scales — Not Just Survives
Most enterprises confuse activity with innovation — running hackathons and pilot programs that never graduate to production. The Innovation Framework dimension gives portfolio leaders a repeatable system for generating, prioritizing, and scaling breakthrough ideas without destabilizing the core delivery engine.
What the Innovation Framework Addresses
Enterprise innovation fails at scale for a predictable set of structural reasons — not because organizations lack creative people, but because the systems surrounding those people are optimised exclusively for predictable delivery. Dimension 02 of the Implementing Safe framework addresses the operating model gap between ideation and industrialisation, embedding structured innovation capacity directly inside the delivery organisation rather than isolating it in a separate innovation lab that the core business ignores.
The framework operates across three horizons simultaneously: sustaining the current value stream, scaling proven experiments into production-grade capabilities, and exploring genuinely disruptive bets. Each horizon demands different governance, funding mechanisms, team configurations, and success metrics. Conflating them — as most scaled agile rollouts do — guarantees that short-term delivery pressure will cannibalize every initiative that does not show a return within the current PI boundary.
The core diagnostic question this dimension answers: does your organisation have a clearly defined, funded, and governed pathway that moves a validated idea from a team’s backlog into a production system and onto a portfolio roadmap — without requiring executive sponsorship at every gate?
How the Framework Is Structured
The Innovation Framework is built around four interconnected mechanisms that together create a closed-loop system: a structured ideation pipeline, a lean validation protocol, a portfolio governance bridge, and a scaling playbook. Each mechanism has defined inputs, owners, cadences, and exit criteria.
The Four Mechanisms
- Ideation Pipeline: A continuously open, team-level submission process tied to strategic themes defined at the portfolio layer. Ideas are tagged by horizon, estimated effort class, and alignment to a named Lean Portfolio Management epic or OKR. This prevents the “ideas in a vacuum” problem where bottom-up creativity has no connection to where the organisation is actually trying to go.
- Lean Validation Protocol: A time-boxed, resource-constrained experimental cycle — typically two to four sprints — with pre-agreed success metrics, a defined pivot/persevere decision gate, and a designated product manager who owns the outcome. No experiment proceeds without a hypothesis statement and a falsifiable success criterion.
- Portfolio Governance Bridge: A quarterly funding review that evaluates validated experiments for graduation into the primary value stream. Experiments that meet threshold criteria receive a named ART allocation and appear on the portfolio Kanban. Those that do not are explicitly closed rather than left in limbo, freeing capacity and reducing organisational debt.
- Scaling Playbook: A documented, repeatable set of steps — covering team formation, architecture review, security and compliance onboarding, and stakeholder alignment — that applies to every initiative graduating from experiment to production. Standardising this path eliminates the ad-hoc negotiation that typically delays scaling by three to six months.
Horizon Allocation Model
Capacity allocation across horizons is explicit and defended at the portfolio level. The model below reflects the starting-point allocation for a mature SAFe organisation. Organisations earlier in their transformation journey typically begin with a smaller Horizon 3 allocation and build toward this target state over six to twelve months.
| Horizon | Focus | Typical Capacity Allocation | Governance Cadence | Primary Success Metric |
|---|---|---|---|---|
| H1 — Sustain | Optimise and protect the current core value stream | 65–70% | PI Planning + Sprint Review | Throughput, defect rate, cycle time |
| H2 — Scale | Grow validated bets into production capabilities | 20–25% | Quarterly portfolio review | Adoption rate, revenue attribution, NPS |
| H3 — Explore | Test disruptive hypotheses with lean experiments | 10–15% | Monthly experiment review | Validated learning per sprint, pivot rate |
Measurable Outcomes This Dimension Drives
Implementing Dimension 02 produces outcomes that are measurable at both the team and portfolio level within two to three PI cycles. The following indicators are tracked in the Implementing Safe maturity dashboard and serve as the primary evidence base for steering investment decisions at portfolio reviews.
- Experiment graduation rate: The percentage of validated experiments that successfully progress to an H2 scaling allocation within one portfolio cycle. Baseline for most organisations entering this dimension sits below 12%; mature implementations consistently reach 35–45%.
- Time-to-hypothesis validation: The average elapsed time from experiment kick-off to a formally recorded pivot-or-persevere decision. Eliminating approval bottlenecks and pre-allocating validation capacity typically reduces this from 14–18 weeks to 4–6 weeks.
- Innovation portfolio coverage: The proportion of active portfolio epics that originated as structured H3 experiments rather than top-down mandates or reactive competitive responses. Higher coverage correlates with stronger alignment between capability investment and actual market signals.
- Capacity leakage reduction: The amount of team capacity previously consumed by informal, ungoverned exploration work — “shadow innovation” — that is recaptured into the structured framework. This is consistently one of the highest-value operational gains in the first six months.
- Strategic theme alignment score: A quarterly assessment of whether the active experiment portfolio maps to the top three to five strategic themes defined at the enterprise level. Misalignment above 40% is a leading indicator of future portfolio debt.
Where teams get stuck
The Experiment Graveyard: Why Good Ideas Never Graduate
- ▸ Experiments are funded ad hoc from team slack time rather than a dedicated, protected capacity allocation — so the moment delivery pressure rises, exploration work stops, and no institutional learning is preserved.
- ▸ Validation criteria are defined retrospectively, after the experiment is already underway, which means pivot-or-persevere decisions are made on sentiment rather than evidence and are reliably challenged by whoever is attached to the original idea.
- ▸ There is no formal hand-off protocol between the experiment team and the ART that would receive the scaled capability, so graduating experiments stall in a no-man’s land of informal negotiation for months — long enough for organisational momentum to evaporate entirely.
Implementation Principles and Anti-Patterns
Effective implementation of the Innovation Framework requires deliberate choices about what not to do as much as what to build. The following anti-patterns appear consistently across enterprise transformations and represent the highest-probability failure modes for this dimension.
Anti-Patterns to Eliminate
- Innovation theatre: Running high-visibility events — hackathons, innovation days, design sprints — that generate energy but have no structural connection to portfolio funding or ART backlogs. The output lives in slide decks. The capacity spent is real.
- Single-horizon thinking: Treating all innovation work as H1 optimisation, measuring every initiative against the same delivery KPIs used for sustaining work. This selectively eliminates the initiatives most likely to produce discontinuous value.
- Centralised innovation gatekeeping: Routing all experiments through a central innovation function that controls funding, tooling, and approval. This creates a bottleneck that slows learning cycles and removes context from the teams closest to the customer problem.
- Undefined pivot criteria: Allowing experiments to continue indefinitely because no one defined what failure looks like before the work started. Zombie experiments consume capacity, demoralise teams, and obscure real portfolio signals.
Enabling Conditions for Dimension 02 Success
| Enabling Condition | Why It Matters | Common Failure Mode Without It |
|---|---|---|
| Protected H3 capacity in the PI plan | Prevents delivery pressure from eliminating exploration work mid-increment | Innovation work is the first casualty of every scope change or production incident |
| Pre-defined pivot-or-persevere criteria | Makes exit decisions evidence-based and defensible at the portfolio level | Sunk-cost bias sustains failing experiments; successful ones are not recognised |
| Named LPM owner for each experiment | Creates accountability for the graduation pathway, not just experiment execution | Experiments succeed technically and then stall because no one owns the next step |
| Documented scaling playbook | Removes the ad-hoc negotiation phase that delays every graduation by months | Each scaling effort reinvents compliance, architecture, and team onboarding from scratch |
| Quarterly innovation portfolio review | Maintains strategic alignment and prevents horizon drift over time | H3 experiments gradually drift toward H1 optimisation as teams seek visible wins |