Martin Eriksson's five-layer model of organizational clarity, built around five questions every organization must be able to answer: where are we going (vision), how are we going to get there (strategy), what's important right now and how do we measure progress (objectives and key results), what actions will we take (opportunities), and how do we choose between those actions (principles).
Each layer answers one question; reading top-to-bottom answers "how", reading bottom-to-top answers "why" (see Laddering (Strategy-Alignment Diagnostic)).
Vision — "Where are we going?" The top layer. For an established business it can sit years out and may never be fully achieved (Google's "organize the world's information" is the standing example). Every team's own vision should ladder up into one shared organizational vision rather than drifting into slightly different versions. See Product Vision.
Strategy — "How are we going to get there?" The most important layer and the one most often missing or done poorly, because it requires committing to trade-offs rather than preserving optionality (see Optionality Avoidance as a Strategy Failure Mode). Set as big bets — which markets, positioning, how to win — top-down from leadership. Explicitly includes "the core to protect": maintaining existing offerings, not just chasing new bets (see Three Buckets of Product Work: New, Existing, Protect Value). Review at least quarterly, not just at the 1–3 year mark it traditionally got (see Continuous Strategy Feedback Loop).
Objectives and Key Results — "What's important right now, and how do we measure progress?" OKRs are one option for this layer, not mandatory — NCTs (Narratives, Commitments, Tasks) and VMOs are named alternatives. Adopt whichever term only after the team agrees what it concretely means, and only if the objectives connect up to strategy and down to opportunities. See OKRs/KPIs as Business-Strategy Mechanism.
Opportunities — "What actions are we actually going to take?" Opportunities that don't connect back to stated strategy should be dropped; the video's own example narrows 10 candidate opportunities down to 4 once the strategy filter is applied. Discovery and experimentation focus only on what survives the filter.
Principles — "How do we choose between those actions?" The foundation layer: a strong reflection of strategy, articulated as a concrete trade-off. Principles arise top-down from well-articulated strategy, or bottom-up when a team hits the same decision repeatedly — escalate it, debate it with data, then codify the resolution instead of re-deciding from scratch each time. Monster's CEO ruling to build for job seekers first, ahead of employers, is the cited example.
The layers work like interlocking blocks: each higher-level decision narrows the options available below it. A clear vision and strategy let teams filter opportunities without re-litigating first principles every time — this is what real delegation depends on.
Eriksson avoids calling the stack a framework so it isn't applied rigidly. Any existing tool — OKRs, NCTs, VMOs — can fill a layer, as long as it connects to the layer above and below. Established organizations shouldn't discard their existing strategy artifacts; the stack is meant to be used diagnostically, to find what's already there, connect it, and refine it — not to replace it wholesale.
The stack is fractal: individual teams have their own 'mini stacks' that should ladder up into the company's 'macro stack.' Teams should only split into their own sub-stack where their answers to the five questions genuinely diverge from the parent stack — not by default.
Communicating the stack works best as continuous co-creation woven into existing ceremonies (all-hands, standups, planning, quarterly kickoffs), not a one-off strategy-offsite reveal. The two most common failure modes in adopting it: trying to build the whole stack at once instead of starting with one layer, and skipping the step of first getting organizational agreement that clarity is actually missing.
Diagnostic techniques for finding where the stack is disconnected: Laddering (Strategy-Alignment Diagnostic), Exec Strategy Interview Diagnostic, Straw Man Approach (Surfacing Strategy Gaps Without Authority), and trade-off surfacing games like This-or-That Trade-Off Game.
Related: Business Strategy, Product Strategy, Mission → Vision → Strategy → Roadmap Hierarchy, Three Buckets of Product Work: New, Existing, Protect Value, Top-Down Bets, Bottom-Up Validation, Continuous Strategy Feedback Loop.
The stack works as an audit tool for a company that already has strategy artifacts, not just a blank-page planning exercise: pull in old documents (e.g. an unused strategy doc from a prior offsite), check whether they still connect to current objectives, and only then challenge or refine them.
Company values commonly fail as decision tools because they don't encode a trade-off — 'we value innovation' or 'we value the customer' doesn't tell anyone what to do when two good things conflict. A principle only becomes useful once stated as a concrete choice: Monster's CEO resolved a jobs-marketplace conflict by ruling 'if we build the best possible experience for job seekers, then recruiters have to follow' — a specific bet, not an aspiration.
Eriksson places principles at the bottom of the stack (below vision, strategy, OKRs, opportunities) rather than near the top next to strategy, even though principles are described as a strong reflection of strategy. This is a deliberate design choice: proposing or challenging a principle feels psychologically safer for a team than challenging strategy directly, so putting principles at ground level where day-to-day work happens makes them more likely to actually get proposed, tested, and revised.
Introducing the decision stack should follow 'think big, but start small': don't try to build all five layers (vision, strategy, OKRs, opportunities, principles) at once, e.g. in a single offsite. Instead, identify whichever single layer is currently weakest — often objectives/OKRs — improve just that layer, verify it ladders correctly to the layers above and below it, and expand outward from there. This mirrors Introducing Focus as a Reversible Test and Single Pilot Team as Transformation Proof of Concept as change-management patterns that de-risk a big structural change by shipping it incrementally rather than as one large reveal.
The oft-cited HBR statistic that 95% of employees can't name their organization's strategy is framed not as evidence that leadership never set a strategy, but that the decision stack decays after being built well once — as the company, team, and market change, the stack silently goes stale unless someone deliberately maintains it. This reframes strategy clarity as a Continuous Strategy Feedback Loop problem rather than a one-time communication or planning failure, and motivates diagnostics like the Exec Strategy Interview Diagnostic as a periodic health check rather than a launch-day exercise.