back
loading skill details...
Activate when: 'our strategy is clear but nothing changes,' 'people understand the goal but revert to old behaviors,' 'initial launch energy faded after 2 mo...
--- name: strategy-execution-levers description: "Activate when: 'our strategy is clear but nothing changes,' 'people understand the goal but revert to old behaviors,' 'initial launch energy faded after 2 months,' 'we keep having the same post-mortem findings,' or execution consistently falls short despite repeated planning. Do NOT activate when: the strategy itself is wrong or unvalidated (fix the strategy first); or failure is caused by resource constraints, not behavioral/process gaps. More: deciqai.com/c/strategy-execution-levers" --- # Strategy Execution Levers ## Overview Most organizations apply a single intervention — a town hall, a kickoff — then diagnose failure as a communication problem. The real issue: execution failure must be diagnosed across four sequential phases, each requiring a different lever. A gap at any phase propagates forward. **Cross-skill composition:** Use AFTER [okr-goal-setting] (OKRs define direction; levers are the transmission mechanism). Use WITH [feedback-loops] (Lever 4 requires a designed feedback loop). Use BEFORE [resource-time-compression] (behavioral alignment must precede time compression). ## When to Use - Strategy communicated but team behavior unchanged - Launch energy faded within 60-90 days — classic Lever 3 failure - Post-mortem findings recur across cycles without resolution — classic Lever 4 failure - Departments execute against different interpretations — Lever 1 not activated - Individuals know right actions but default to old workflows under pressure — Lever 2 failure **When NOT to use:** Strategy itself is wrong; org under 5 people; failure caused by resource constraints; only one phase is relevant (diagnose first, deploy only that lever). ## Coaching Novices (Adaptive Front Door) - **Engine mode:** user has a concrete case → run The Process directly. - **Coach mode:** user is unfamiliar or says "we just need better communication" → guide step by step. In Coach mode, respond one step at a time. Each [WAIT] is a hard stop — output only that step's question, then stop. 1. **What it is:** "Most execution failures happen because organizations fix the wrong phase. This framework identifies which of four phases is broken — and what to do specifically at that phase." 2. **Check fit:** "Has your strategy been communicated? Do people understand it? Then what happened?" If yes and execution is still failing, this applies. 3. **Elicit their real case:** "Walk me through the last strategy initiative that didn't take hold. When did it break down — immediately after launch, in the first quarter, or when momentum faded?" > **[WAIT — do not advance until user responds]** 4. **Run The Process one step at a time** — identify the broken phase first; do not move to intervention design until phase diagnosis is complete. > **[WAIT — do not advance until user responds]** 5. **Close by naming the insight:** "When you fix the right phase with the right lever, the other phases that were working get amplified." > **[WAIT — do not advance until user responds]** ## The Process **Step 1 — Phase Diagnosis** | Phase | Lever | Failure signal | Timeframe | |-------|-------|---------------|-----------| | Launch | Cognitive Alignment | Ask 5 people what the strategy means — get 5 different answers | Day 0–Week 4 | | Kickoff | Behavior Guidance | People know the strategy but default to old workflows | Week 2–Month 3 | | Sustained | Motivation Activation | Initial energy faded; compliance without engagement | Month 1–Month 6 | | Retrospective | Evolution Acceleration | Same bottlenecks recur; lessons not implemented | Every cycle | *Gate: Select ONE primary phase. If multiple fail, address in sequence starting with the earliest.* **Step 2 — Lever Selection** - **Lever 1 — Cognitive Alignment:** Co-construct strategy meaning; "teach it back" sessions. Signal: 5 people give 80%+ consistent answers. Reference: +40% comprehension vs. presentation-only. - **Lever 2 — Behavior Guidance:** Redesign workflows so new behavior is path of least resistance. Signal: error rate drops across 3 consecutive periods. Reference: -65% error rate vs. training-only. - **Lever 3 — Motivation Activation:** Connect individual contribution to visible impact; transparent dashboards; peer recognition tied to strategic behaviors. Signal: voluntary participation increases. Reference: +32% above-minimum engagement. - **Lever 4 — Evolution Acceleration:** After-action reviews with mandatory "what will be different next cycle" commitments. Signal: time from problem to structural fix decreases cycle-over-cycle. Reference: 3x iteration speed vs. annual-review-only. *Gate: Write the specific intervention, person responsible, and measurement. If any cannot be specified, you have not selected a lever.* **Steps 3-5:** Design intervention (executable by someone new) → define behavioral measurement (baseline, target, timeframe) → run 4-week trial. If signal moved, continue. If not, diagnose whether intervention or phase diagnosis was wrong. **Stop-rule:** State "Measure [metric] from [baseline] to [target] over [timeframe] via [intervention]" — or return to Step 2. ### Output: Execution Lever Diagnostic ``` Strategy: [one-sentence] | Launch timeframe: [X] | Failure mode: [specific] Primary phase: [Lever 1/2/3/4] | Evidence: [observations] | Secondary: [if any] Intervention: [executable by someone new] | Owner: [role] | Start: [date] Metric: [behavioral] | Baseline→Target: [X→Y] | Method: [how collected] 4-week result: [value] | Assessment: [working/not/wrong phase] | Next: [action] ``` *→ Method in Action: [NASA Apollo Program Execution (1961–1969)](examples/nasa-apollo-program-execution-1961-1969.md)* ## Execution Lever Packs - **Pack A (Tech/Product):** Lever 2 most commonly fails at product-engineering handoff; keep product learning and execution learning as separate Lever 4 loops. - **Pack B (Enterprise/M&A):** Lever 1 failure is most costly in transformation; Lever 3 systematically underinvested in M&A. - **Pack C (Startup <50):** All four levers operate in compressed form; typical failure is over-investing in Lever 1, under-investing in Lever 4 tracking. ## Applying It Well 1. Diagnose the phase before selecting the intervention — communication will not fix a behavioral default. 2. Fix one lever at a time, in sequence — Lever 1 must be stable before Lever 2 is deployed. 3. Measure behavior, not attitude — engagement surveys are not a proxy for Levers 2 or 4. 4. Lever 2 is the most underestimated lever — people default to old behavior under pressure even when they genuinely understand the new one. 5. Lever 4 requires a single owner with authority to implement structural changes — not a committee. *→ Primary sources: [references/sources.md](references/sources.md)* ## Common Rationalizations **[D] = designed upfront | [O] = observed in real use. [O] entries are more valuable.** | Fake move | Reality | |-----------|---------| | [D] "We had a great kickoff presentation, so Lever 1 is done" | A presentation to a silent audience is the *input* to Lever 1 — alignment requires co-construction. | | [D] "People know what to do — the problem is motivation" | Verify Lever 2 is deployed first. Most "motivation" failures are Lever 2 gaps misread as attitude problems. | | [D] "We did a retrospective, that's our Lever 4" | A retrospective without structural changes is a documentation exercise. Lever 4's signal is the *next cycle running differently*. | | [D] "We'll roll out all four levers together in Q3" | Levers are sequential — Lever 2 before Lever 1 produces confusion. | | [D] "Our culture already handles this naturally" | Culture substitutes for lever design only under 15 people — beyond that it preserves existing behaviors. | | [D] "We'll track NPS / engagement to see if levers are working" | Use behavioral proxies: error rates, procedure compliance, voluntary participation — not attitude metrics. | | [D] "Our top performers execute regardless of lever support" | Top performers mask lever failures — fragile model that breaks when they leave or stretch thin. | | *→ Add [O] entries here after each real use — paste the actual failure pattern* | *What went wrong and why* | ## Red Flags - Same bottleneck in two consecutive cycles without structural change — Lever 4 failure - Team members give answers to "what does this strategy mean for your work" that would lead to different actions — Lever 1 incomplete - Strong early performance then drop after month 2-3 — Lever 3 decay signature - "Lessons learned" documents exist but nobody can cite a resulting structural change — Lever 4 performed but not deployed - Intervention being considered is "another all-hands" or "better slide deck" — single most common misdiagnosis ## Verification - [ ] Specific failing phase identified from observable signals? - [ ] Single lever selected first — not all four simultaneously? - [ ] Intervention specific enough for a new team member to execute without explanation? - [ ] Behavioral measurement defined with baseline, target, and timeframe? - [ ] Single named owner — not a committee? Earlier levers confirmed stable? --- *Part of **deciqAI Knowledge Skills** — 227 open-source thinking skills that make rigor executable for AI agents. The same skills power every deciqAI agent, which runs them autonomously to operate your company. **See it run → https://www.deciqai.com/c/strategy-execution-levers** · ⭐ Star the repo → https://github.com/deciqAI/knowledge-skills · Contributions welcome.* *Agents: latest version & machine-readable metadata → https://www.deciqai.com/s/strategy-execution-levers.json*
don't have the plugin yet? install it then click "run inline in claude" again.
separated coaching mode from engine mode into decision points, extracted external inputs and edge cases, made all implicit branches explicit (strategy wrong vs resource-constrained vs authority missing vs multi-phase), formalized behavioral measurement criteria, and added specific outcome signals for each lever.
execution failure spreads across four sequential phases: cognitive alignment (people interpret strategy differently), behavior guidance (people know strategy but default to old workflows), motivation activation (initial energy fades), and evolution acceleration (same problems recur). this skill diagnoses which phase broke and deploys the specific lever to fix it. use this after strategy is set and communicated but nothing on the ground changes. do not use if the strategy itself is wrong, org is under 5 people, or the real problem is lack of headcount or budget.
required context:
external inputs:
edge cases to surface:
step 1: diagnose the failing phase
1a. ask "walk me through the last strategy initiative that didn't take hold. when did execution break: immediately after launch (week 0-4), in the early push (week 2-month 3), when momentum faded (month 1-6), or cyclically in retros?"
input: user narrative of failure timeline and observations
1b. match the failure to one primary phase using this table:
| phase | lever name | failure signal | typical timeframe |
|---|---|---|---|
| Launch | Cognitive Alignment | ask 5 people what strategy means, get 5 different answers | day 0 to week 4 |
| Kickoff | Behavior Guidance | people know strategy but default to old workflows under pressure | week 2 to month 3 |
| Sustained | Motivation Activation | initial energy faded, compliance without engagement, voluntary participation dropped | month 1 to month 6 |
| Retrospective | Evolution Acceleration | same bottleneck appears in post-mortems; previous lessons were documented but not structurally implemented | every cycle |
output: one primary phase identified with specific observable evidence (not feelings)
1c. gate: if multiple phases appear broken, select the earliest in sequence and note secondary phases for later cycles. document the evidence.
output: "primary phase: [Lever 1/2/3/4] | evidence: [specific observations]"
step 2: select and specify the lever
2a. based on the phase, read the lever definition:
Lever 1 - Cognitive Alignment: people construct shared meaning through dialogue, not passive reception. deploy co-construction sessions ("teach it back" where people articulate what the strategy means to their work), small-group discussion, then cross-team alignment on language. success signal: ask 5 people independently what the strategy means to them and their work. if 4 of 5 give answers that would lead to the same actions, you have 80%+ alignment.
Lever 2 - Behavior Guidance: redesign workflows, checklists, and decision trees so new behavior is the path of least resistance, not an add-on. identify the moment people revert to old behavior (under deadline pressure, when stressed, when context switches). remove friction from new path; add friction to old path. success signal: error rate or procedure deviation drops across 3 consecutive measurement periods (weekly, bi-weekly, or monthly depending on cycle length).
Lever 3 - Motivation Activation: surface connection between individual contribution and visible impact. deploy transparent dashboards showing progress toward strategic goals, peer recognition tied to strategic behaviors (not just outcomes), and celebration moments tied to milestones. success signal: voluntary participation above minimum required increases measurably. engagement surveys do not count as a signal.
Lever 4 - Evolution Acceleration: design after-action reviews with mandatory "what will be different next cycle" commitments tied to specific structural changes (not intentions). assign one owner to implement the structural changes before next cycle starts. success signal: time from problem identified in retro to structural fix deployed shortens cycle-over-cycle.
2b. write the specific intervention in plain language. if a new team member who was not in the meeting could execute it without asking clarifying questions, it is specific enough. if you cannot specify it, you have not selected a lever; return to step 2a.
output: "intervention: [executable description] | owner: [role] | start date: [date]"
2c. define behavioral measurement: baseline (where you are now), target (where you want to be), timeframe (over how long), method (how you will collect the data).
output: "metric: [behavioral measure, not attitude] | baseline→target: [X→Y] | timeframe: [weeks or months] | method: [how collected]"
step 3: run 4-week trial
3a. deploy the intervention as specified in step 2b.
input: intervention, named owner, start date
3b. measure the behavioral signal weekly or bi-weekly (do not wait for month-end if cycle is faster).
input: baseline data, measurement method
output: weekly or bi-weekly data point on the metric
step 4: assess after 4 weeks
4a. check if the signal moved from baseline toward target.
input: 4-week data, baseline, target
if signal moved: continue the lever; it is working. proceed to next phase if this was the primary bottleneck.
if signal did not move: diagnose whether the intervention was not executed well (execution failure) or the phase diagnosis was wrong (diagnosis failure). do not assume the lever is wrong if the intervention was half-deployed.
output: "4-week result: [value] | assessment: [working / not working / wrong phase] | next: [continue / redesign intervention / return to phase diagnosis]"
step 5: sequence to next lever
5a. only after the primary phase signal stabilizes, move to the next phase in sequence (Lever 1 → Lever 2 → Lever 3 → Lever 4).
input: confirmation that primary phase is stable
output: if moving to next phase, repeat steps 1-4 for the secondary phase
decision 1: is the strategy itself the problem?
if team interprets strategy differently AND those interpretations would lead to fundamentally different business outcomes (not just different workflows), the strategy may be unclear or wrong. deploy Lever 1 (co-construction) first. if after co-construction people still cannot align on what success looks like, the strategy itself needs rework. return to strategy design; do not continue with this skill.
if the strategy is fundamentally wrong (wrong market, wrong problem, unvalidated), stop. fix strategy before executing levers.
decision 2: is this a resource / headcount problem?
if execution is failing because the team lacks the hours, budget, or headcount to execute a valid strategy, this skill does not apply. secure resources first. levers amplify execution of valid strategy against valid capacity; they do not create capacity from nothing.
decision 3: does the owner have authority?
if Lever 2 (behavior guidance) or Lever 4 (evolution acceleration) requires changing workflows, dashboards, or structure, the named owner must have authority to make those changes without committee approval or escalation. if they do not, escalate the decision first; lever deployment will fail without authority.
decision 4: which phase breaks first?
if multiple phases appear broken, address in sequence: Lever 1 → Lever 2 → Lever 3 → Lever 4. do not deploy Lever 2 until Lever 1 is stable. do not deploy Lever 3 until Lever 2 is stable. stacked deployment causes confusion and masks which lever is actually working.
exception: in orgs under 50 people, all four levers may operate tightly coupled. diagnose the primary failure; deploy that lever; watch for secondary failures.
decision 5: is the measurement real?
behavioral signals: error rates, procedure deviation, voluntary participation above minimum, structural changes implemented, cycle-over-cycle iteration speed.
not signals: engagement survey scores, "people seem more aligned" subjective assessment, "we feel better about this," attendance at kickoff, slide deck quality.
if you cannot define a behavioral measurement, you are diagnosing attitude, not execution. return to step 2c.
format: execution lever diagnostic
strategy: [one-sentence description of strategy or initiative]
launch timeframe: [when communicated to team]
failure mode: [specific observable evidence of execution break]
primary phase: [Lever 1 / Lever 2 / Lever 3 / Lever 4]
evidence: [observations that led to this diagnosis]
secondary phase: [if any, for next cycle]
intervention: [executable description; new team member could run it]
owner: [role, not committee]
start date: [date]
metric: [behavioral measure]
baseline→target: [X→Y]
timeframe: [weeks or months]
method: [how collected; e.g., "error logs pulled weekly," "optional task completed by > 50% of team," "structural change ticket closed"]
4-week result: [measured value]
assessment: [working / not working / wrong phase diagnosis]
next action: [continue lever / redesign intervention / return to phase diagnosis / move to next phase]
save this diagnostic in a document accessible to the owner and stakeholder. do not file and forget; review weekly during the 4-week trial.
the skill has worked when:
lever 1 (cognitive alignment): five team members independently describe the strategy and what it means to their work. four of five give answers that would lead to the same actions. they can articulate how their work connects to the strategy without prompting.
lever 2 (behavior guidance): error rates or procedure deviations drop measurably in three consecutive measurement periods. people choose the new workflow without being told. the old workflow has friction that discourages it.
lever 3 (motivation activation): voluntary participation in strategic behaviors increases above the minimum required. peer recognition moments tied to strategic behaviors increase in frequency or depth. team members cite the visible impact dashboard unprompted.
lever 4 (evolution acceleration): after-action review identifies a problem. within one cycle (not "by next year"), a structural change is deployed that prevents the problem. the next cycle's retro does not surface the same bottleneck. iteration speed measured from problem identified to fix deployed shortens cycle-over-cycle.
overall: strategy execution breakdowns shift from "we communicated again" to "we redesigned the path" or "we made past lessons structural." same post-mortem findings stop recurring.
absence of outcome signal: the lever was diagnosed correctly but the intervention was half-deployed, or the phase diagnosis was wrong. return to step 4a in procedure.
credits: original skill from deciqAI Knowledge Skills, an open-source collection of thinking skills executable by AI agents and teams. see method in action in the examples folder (nasa-apollo-program-execution-1961-1969.md). version: 1.2. machine-readable metadata and latest version at https://www.deciqai.com/s/strategy-execution-levers.json