Improve Business

Improve a Business

An interactive, resumable journey for a business that already runs but has stopped moving — real revenue, real customers, a flat line. Across eight phases it works the layers in order (strategy, execution, management, structure, motivation, metrics, margin, pricing), diagnosing before treating rather than chasing one tactic at a time. The agent asks before every decision and records each choice in docs/ so the turnaround survives across sessions.

Core Principle

A plateau is a diagnosis failure: diagnose before treating, and work the layers in order — strategy, execution, management, structure, motivation, metrics, margin, pricing. Poking one tactic at a time never breaks a plateau because the constraints compound. This skill sequences the phases, asks the decision questions, and records what was decided; the constituent skills carry the actual method — invoke them rather than improvising their frameworks.

Journey Map

Operating Rules

1. Resume first. Before anything else, read docs/IMPROVE-BUSINESS-PLAN.md and every artifact in the Journey Map. If the tracker exists, summarize the journey state in 3-5 lines and ask which phase to enter. Done when the user has confirmed an entry point. A journey with a tracker is resumed, never restarted.
2. Intake on first run only. No tracker: run the Intake below, then create docs/IMPROVE-BUSINESS-PLAN.md with every phase statused pending | in-progress | awaiting-evidence | done | deferred: reason | skipped: reason. Done when the tracker exists and the user has confirmed the phase plan.
3. Phase entry. Announce: what the phase does, the decision it forces, the artifact it produces, rough effort. Offer proceed / skip / defer — phases marked GATE may be deferred, never skipped. Mark the phase in-progress on proceed. Done when the user chose.
4. Skill invocation and fallback. Load the phase's skill and use it: each phase's Invoke line names the skill by slug — use that skill to run the phase. If it is not available, offer: npx skills add wondelai/skills/<slug> --global. If the user declines, run the phase from its Brief — the minimum viable method. State which mode you are in.
5. In-phase decisions. Ask every question under "Decide with the user" — with concrete options and your recommendation. Record the choice in the tracker's Key Decisions. A decision made silently is a defect.
6. Phase exit. Present the draft artifact content for sign-off before writing. On approval: write or extend the docs/ files, update the tracker (status, Key Decisions, Next Actions). Done when the files are written and the phase row shows done.
7. Artifact discipline. Read before writing; create a file only if missing, otherwise extend — add or update your sections, preserve everyone else's. Files are UPPERCASE in docs/. Every recommendation lands as a checkbox or a table row with owner and priority. See references/artifact-templates.md when creating a docs/ file for the first time — create it from the full skeleton (all section headings), then fill the sections your phase names.
8. Diagnosis before treatment. No new initiative, reorg, or pricing change starts before docs/STRATEGY.md holds a written Diagnosis the user has confirmed. When asked to jump straight to Phase 4, 7, or 8, complete Phase 1 first (or confirm it is already done and recorded). Done when a confirmed Diagnosis exists before any treatment phase writes its artifact.

Intake

Ask these before creating the tracker:

1. What is the plateau — flat revenue, rising churn, a team working hard with nothing moving? (Frames the Phase 1 diagnosis and the Phase 6 OMTM.)
2. What passes for your current strategy today — a deck, OKRs, a goal list, nothing written? (Feeds Phase 1; if none exists the diagnosis work is larger.)
3. How big is the team and how is it structured? (Sizes the Rocks and 1:1 system in Phases 2-3; gates Phase 4.)
4. Do you run technology teams, or is this a services / brick-and-mortar business? (team-topologies is engineering-specific — skip Phase 4 for non-tech orgs.)
5. When was pricing last set, and on what basis — cost-plus, competitor, or value? (Gates Phase 8; cost/competitor pricing signals margin left on the table.)
6. What is your sales win rate, and do you hear price objections? (Near-100% win rate with no pushback signals minivation — Phase 8.)
7. Where does work get stuck — within teams or between them? (Between-team stalls point at Phase 4; within-team points at Phases 3 and 5.)

Phase-skip heuristics: skip Phase 4 when you don't run technology teams or teams already deliver cleanly; skip Phase 5 when morale is engaged and not the constraint; skip Phase 7 when no large renewals or vendor contracts are in play; skip Phase 8 only if pricing was recently repriced from validated willingness to pay.

Then create docs/IMPROVE-BUSINESS-PLAN.md from the template and confirm the phase plan. Done when docs/IMPROVE-BUSINESS-PLAN.md exists with every phase statused and the user has confirmed the plan.

Phases

Phase 1 — Diagnose the real challenge (good-strategy-bad-strategy) — GATE

Purpose: Name the single critical challenge capping the business, so every later layer works the right constraint. Phases 2-8 may not start before this Diagnosis is confirmed.

Brief (fallback): Good strategy has a kernel — an honest diagnosis of the critical challenge, a guiding policy, and coherent actions. Bad strategy is fluff, goals dressed as strategy, and forty "priorities" that concentrate nothing. Treat the funnel as a chain-link system and find the one binding link; investing in strong links is wasted until the weak one is fixed. Set a proximate objective close enough that the team can see how to hit it, and force a no-list.

Invoke: Use the good-strategy-bad-strategy skill with the current annual plan, OKRs, or deck pasted in. Ask it to (a) audit the plan as strategy and score it 0-10, (b) write a proper kernel, (c) name the binding chain-link constraint and one proximate objective for the quarter.

Decide with the user: Does the diagnosis name the real obstacle, not a symptom? Which single link (lead-gen / activation / retention / expansion) is the binding constraint, and what is the proximate objective? What goes on the no-list — what will you deliberately stop doing?

Artifact: Create docs/STRATEGY.md with ## Diagnosis (critical challenge + binding constraint), ## Guiding Policy (including the proximate objective), ## Coherent Actions (Action | Owner | Due | Status), and ## No-List. Update the tracker.

Done when: STRATEGY.md holds a confirmed Diagnosis, a Guiding Policy, 3-5 Coherent Actions with owners, and a No-List — only then are Phases 2-8 unlocked.

Phase 2 — Install an execution rhythm (traction-eos)

Purpose: Convert the strategy into a weekly operating rhythm that survives Monday.

Brief (fallback): Vision without traction is hallucination. Set 3-7 quarterly Rocks — each with one owner, SMART, binary (done or not, no partial credit) — aimed at the proximate objective. Install the Level 10 Meeting: same day and time, 90 minutes, fixed agenda ending in 60 minutes of IDS (Identify, Discuss, Solve) driven to root cause. Never cancel it.

Invoke: Use the traction-eos skill with STRATEGY.md's guiding policy and proximate objective. Ask for (a) 5-7 quarterly Rocks with owners and binary criteria, flagging any that are business-as-usual, (b) a Level 10 agenda with time boxes, (c) an IDS pass on the top recurring issue.

Decide with the user: Which 5-7 Rocks make the quarter (cap at seven)? Who owns each? What weekly cadence for the Level 10, and who facilitates?

Artifact: Create docs/OPERATIONS.md with ## Vision/Traction Summary, ## Rocks (this quarter) (Rock | Owner | Binary done-condition | Status), and ## Meeting Cadence (Meeting | Rhythm | Agenda). Update the tracker.

Done when: OPERATIONS.md holds 5-7 owned binary Rocks tied to the proximate objective and a scheduled Level 10 cadence the user has committed to.

Phase 3 — Manage for output, not activity (high-output-management)

Purpose: Multiply the team's output by getting the operator off the critical path.

Brief (fallback): A manager's output equals the output of their organization plus the neighboring teams they influence. Negative-leverage habits — meddling, waffling, being the approval bottleneck — multiply downward. High-leverage tools: subordinate-owned 1:1s at a cadence set by task-relevant maturity (TRM), and delegation-with-monitoring — delegate what you know best, monitor at task level, sample deeper when TRM is low, lighter as it rises.

Invoke: Use the high-output-management skill with the operator's last two weeks of calendar and the tasks they keep grabbing back. Ask for a leverage audit, a redesigned week built around forecasted key events, and a 1:1 plus delegation system for direct reports.

Decide with the user: Which negative-leverage habits to cut? The 1:1 cadence per report (set by TRM, not seniority)? Which tasks to delegate-with-monitoring off the critical path?

Artifact: Extend docs/OPERATIONS.md: add ## Management Leverage (1:1 cadence per report, delegation decisions, monitoring depth). Update the tracker.

Done when: OPERATIONS.md records a 1:1 cadence for every direct report and at least one delegation decision that removes the operator as a bottleneck.

Phase 4 — Reshape teams around value streams (team-topologies)

Purpose: Remove structural friction when work keeps getting stuck between teams rather than within them.

Brief (fallback): Conway's law — you ship your communication structure. Reduce every team to one of four types: stream-aligned (the default), enabling, complicated-subsystem, platform. Cap cognitive load: at most one complicated domain per team; never split a single domain across two teams. Shed domains or remove extraneous load before adding headcount. Define each interaction mode: collaboration, X-as-a-service, or facilitating.

Invoke: Use the team-topologies skill with each current team and what it does day to day. Ask it to classify teams, flag anti-patterns (shared-services ticket queue, teams split by function not value stream), diagnose cognitive load, and propose value-stream-aligned boundaries with explicit interaction modes.

Decide with the user: First — is the plateau structural (handoffs dominate lead time, recurring coordination meetings, everyone's-and-no-one's ownership)? If not, skip. If yes: which domains to shed, which extraneous load to platform-ize.

Artifact: Extend docs/OPERATIONS.md: add ## Team Structure (team topology, per-team type, cognitive-load notes, interaction modes). Update the tracker.

Done when: either the phase is skipped with a recorded reason, or OPERATIONS.md holds a four-type classification and a proposed value-stream structure with interaction modes.