SuperAPI AI Directory

marketing-plan-orchestrator

Marketing & growth

View repository

Updated 10-16 06:13

About this skill

Marketing Plan Orchestrator

Turn customer evidence and commercial constraints into a complete, standalone marketing plan: what market to serve, what to say, where and how often to say it, how demand becomes cash, what it costs, who does what, and how the plan is steered. A marketing plan is a distinct deliverable; it must work without a full business plan, and every objective must be specific, measurable, accurate, realistic, time-bound and applicable to the business, its location and scope.

Use When

  • Create or substantially revise an annual, launch, product, service, market-entry, regional or campaign marketing plan.
  • Audit an existing marketing plan against the document architecture and quality gate.
  • Reconcile brand, demand, sales, service, channel investment and financial assumptions across a marketing programme.

Do Not Use When

  • Use 07-marketing-sales-strategy for Section 07 of a business plan; it borrows this skill's references at section depth.
  • Use digital-marketing-strategy for only the digital layer. Use the digital marketing and advertising engine (social-media-skills) for advertising strategy detail, detailed media plans, creative briefs, build specifications, optimisation and attribution; this skill keeps only the plan-level objectives, budget envelope and media-mix decision reconciled to the P&L.
  • Use meta-strategic-factor-analysis or meta-strategic-options-evaluation for corporate strategy work that is not a marketing plan.
  • Do not replace a full business plan, statutory review or evidence of market demand with this workflow.

Required Inputs

Artefact Source/provider Required? Behaviour when missing
Business, offer, location and scope, audience, decision, planning horizon and strategy type Sponsor or 00-client-intake Yes Stop strategic choices; request the missing context
Customer, competitor and current-performance evidence (sales, CRM, research) Research, CRM, sales and service records Yes Produce a research plan and bounded hypotheses; no validated demand claim
Price, contribution, budget, cash and capacity constraints Finance and delivery owners Yes Hold scale and budget approval; use labelled scenarios
Brand, distribution, sales, service and delivery commitments Accountable owners Yes Record dependencies and required decisions
Current legal, platform and market facts Currentness register, Digital Research engine, qualified reviewers Conditional Mark affected claims and release gates NOT_ASSESSED

Workflow

  1. Frame. Record the business outcome, decision, reader, horizon, budget authority, strategy type and scope level. Apply location and scope calibration. Open a claim register using evidence discipline.
  2. Outline. Choose the depth from the scaling rules in document architecture. Keep all 20 section decisions; compress, do not omit.
  3. Research the customer first. Build the customer insights report, segmentation, target-market tests, buying situations and sensitivities using the toolkit and the decision workpapers. Stop creative and media work until this exists.
  4. Analyse the situation. Reduce PESTEL to key drivers, rate five forces, and build EFAS, IFAS and SFAS on the single scale defined by meta-strategic-factor-analysis; convert SWOT to TOWS. Size the market top-down and bottom-up with a sanity check. Map competitors with weighted grids.
  5. Position. Write the positioning statement, choose one or two of the five buyer-benefit answers and two or three message themes by importance × proof × distinctiveness. Route premium offers to meta-premium-go-to-market.
  6. Set objectives. Build the business → marketing → communication → channel cascade with the SMART objective builder. Reject any objective that fails its quality test.
  7. Choose strategy. STP and coverage, strategy-clock position, growth route and method, push and pull balance, and the marketing mix with a decision per P; evaluate options with meta-strategic-options-evaluation when the choice is contested.
  8. Choose channels. Run the Bullseye method across all 19 channels; write three test cards; set the critical path and not-doing list.
  9. Plan advertising, digital and content. Use the advertising and media plan for objectives, triangulated budget, reach, frequency, flighting, the audience-cost table and the creative brief. Hand the digital, content and detailed advertising layers to digital-marketing-strategy and the digital marketing and advertising engine with the handoff table in the architecture; keep the envelope, mix and P&L reconciliation here.
  10. Plan sales and retention. Apply sales plan and account coverage and the retention and CLV sections of marketing economics.
  11. Cost and reconcile. Build the two-phase budget, direct-response break-even lines, the maximum affordable cost per lead and the demand-to-cash bridge; reconcile the budget with the P&L marketing line under the Chwezi finance doctrine. Stop release on any mismatch.
  12. Steer. Build the calendar, KPI dashboard (OMTM, guard-rails, lines in the sand), control table, war-room cadence and risk register with the KPI and control plan.
  13. Write and gate. Draft the executive summary last. Write with the marketing-plan phrase bank, writing-quality and anti-ai-slop. Run the quality gate; when it fails, return to the affected step, revise, and re-run the gate. Hand the plan and evidence to named owners.

Decision Rules

Condition Action Failure or risk avoided
Customer research or an approved plan is missing Hold creative and scaled media; run research and a capped test first Spending on messages nobody has tested
Demand evidence is weak Commission a bounded validation test before scaling A persuasive forecast masks an untested market
An objective fails the SMART quality test or is not calibrated to the place and scope Rewrite it or delete it Unmeasurable or inapplicable targets
More leads would exceed delivery or sales capacity Repair delivery, narrow demand or sequence investment Marketing damages the brand promise
Budget below the minimum effective level Concentrate on fewer places, segments or months Invisible, spread-thin campaigns
A channel's cost per enquiry exceeds the affordable ceiling after its test Drop or redesign it and move to the next test card Scaling a loss-making channel
Premium price has no distinct buyer value or proof Revise offer and test price acceptance Prestige language substitutes for willingness to pay
Digital is not the best route for the segment Prioritise a justified offline, partner or sales route Channel fashion overrides customer behaviour
Plan budget differs from the P&L marketing line Return to the owning section and reconcile before release Narrative and numbers that disagree
Direct marketing or data collection lacks a data-protection line Add registration, consent and objection handling per jurisdiction Unlawful campaigns and regulator exposure
A material source, model or approval is missing Hold the affected decision and name the recovery input False release readiness

Outputs

Artefact Consumer Acceptance condition
Standalone marketing plan (20 sections plus appendices) Sponsor and operating team Passes the quality gate with no blockers and meets the release thresholds
Objectives cascade and KPI table Owners and reviewers Every objective passes the SMART test; every KPI has a line, owner and pre-agreed action
Budget, economics and demand-to-cash model Finance and commercial owners Budget equals the P&L line; costs, demand, margin, cash and capacity reconcile by period and scenario
Channel, media and sales plans Channel, sales and service owners Every initiative has owner, budget, dependency, measure, stop rule and review date
Handoff briefs Digital marketing and advertising, website and design engines Message, channel and compliance briefs carry the agreed definitions
Evidence and release record Reviewer Claims classed and sourced; unknowns and NOT_ASSESSED items listed

Evidence Produced

Evidence Format Acceptance condition
Claim and source register Appendix table Every number has a class, source and date; register claim IDs cited
Customer insights report and research records Workpaper Method, sample, dates and limitations visible; no fabricated respondents
Options and investment decisions Decision table and model A reviewer can reproduce the selected option, rejected alternatives and downside case
Quality-gate record Scored rubric Score, blockers, fixes and reviewer recorded; missing review marked NOT_ASSESSED

Quality Standards

  • Make explicit choices; a list of channels is not a strategy.
  • Every objective is SMART and applicable to the place and scope; every target states the source of its line in the sand.
  • Reconcile the promise with delivery quality, the forecast with cash and capacity, and the budget with the P&L.
  • Separate verified facts, assumptions, estimates, projections and targets; use register or freshly verified current facts only.
  • Never call the plan bankable, viable, investor-ready or compliant without the engine's evidence gates.
  • Review authored prose and rendered artefacts separately. Missing market, finance, professional or render evidence remains NOT_ASSESSED.

Anti-Patterns

  • Delivering only a social calendar. Fix: cover market, offer, price, distribution, sales, economics and control.
  • Treating affluent consumers or "SMEs" as one segment. Fix: apply the four target-market tests and research buying situations.
  • Budgeting media alone. Fix: include production, research, testing, labour, fees, tools and contingency, triangulated and reconciled.
  • Forecasting sales from impressions. Fix: model qualified demand, conversion, sales lag and fulfilment with explicit assumptions.
  • Copying foreign benchmarks and channel habits into a Ugandan plan. Fix: calibrate to the place, scope and evidence.
  • Calling the plan convincing because it is polished. Fix: challenge the evidence, alternatives and downside economics.
  • Applying every framework. Fix: retain only methods that change a decision or expose a risk.

Capability Contract

Read and search supplied evidence; calculate and draft the authorised working plan. Analysis, planning and gate review are read-only with respect to source systems. Network verification is required for current claims. Publishing, outreach, spending, account changes and certification require explicit authority.

Degraded Mode

Without research, financial inputs or current sources, deliver a labelled hypothesis plan with the evidence-acquisition sequence, test cards and placeholders listed as evidence requests. Hold investment approval and affected compliance claims. If export or render tools are unavailable, deliver Markdown and mark final-format checks not assessed; do not claim files exist.

Worked Example

A Kampala meal-prep business wants "a big Facebook campaign". The orchestrator first builds a customer insights report (320 survey responses and 10 interviews in Kira and Nakawa), finds that working parents value "no cooking after work" and weekly mobile-money payment, and that month-two reorder is the weak point. It sets a reconciled cascade (UGX 38m to UGX 55m monthly revenue by June 2027; 160 net new households; month-two reorder 48% to 60%), runs Bullseye tests on click-to-WhatsApp ads, resident-association sponsorships and a referral credit, caps each at UGX 1.5m, and makes month-two reorder the OMTM. The budget equals the P&L marketing line; a third delivery rider is scheduled before capacity binds. All figures illustrative.

References

Read Next

  • meta-strategic-factor-analysis — EFAS, IFAS and SFAS tables for section 2.
  • meta-strategic-options-evaluation — contested strategic choices in section 8.
  • demand-forecasting — volume forecasts behind the demand-to-cash bridge.
  • ai-slop-audit — graded review after each drafted section.