54-media-plan-global

# Media Plan (Global)

A plan without hard numbers cannot tell you whether you are winning. Every figure in the plan must trace back to a revenue target. Run `10-reverse-kpi-global` for max CPA and required budget before allocating anything.

## Information gathering

Read `.agents/product-marketing-context-global.md` plus any output from `51-audience-research-global` and `10-reverse-kpi-global`. If information is missing, ask up to 4 questions:

1. **Which offer, and what is the revenue target?** Target revenue for the period, AOV, and gross margin percentage.
2. **What is the budget?** A hard cap, or a number derived backwards from the revenue target?
3. **Which channels and markets?** Meta / Google Search / TikTok / YouTube / Pinterest / LinkedIn — and which countries. Primary objective?
4. **What is the timeline?** Number of weeks, plus any launch date or seasonal moment that cannot move.

## Principles

1. **Work backwards from revenue, never forwards from "the budget we have".** Revenue -> orders -> leads -> max CPA -> budget. If available budget is below required budget, lower the target or lower the CPA. Do not pretend.
2. **Market tier is a planning input, not a detail.** Per `references/benchmarks-global.md`, Tier 1 (US, Canada, Australia, Western EU) costs 6-7x Tier 2 (SEA, LATAM). A campaign profitable at 3x ROAS in a Tier 2 market may need 5-7x in the US for the same unit economics. Plan and report per market.
3. **Season changes the math.** Q1 is the cheapest window (January median CPM $15.74, about 21% below average). November peaks at $25.22, roughly 27% above average. A Q4 plan built on Q1 costs is already wrong.
4. **Spread across the funnel.** Running only bottom-funnel from day one produces high CPA because no trust exists yet.
5. **No blank KPIs.** Every channel gets its own CPA and efficiency target.
6. **Test first, scale second.** Weeks 1-2 are tuition. Do not judge ROAS in week one.
7. **Review weekly and adjust.** The plan is a hypothesis; live data decides.

## Workflow

### 1. Reverse-calculate the KPIs (run `10-reverse-kpi-global`)

```
Target revenue / AOV        = Orders needed
Orders / close rate         = Leads needed
Max CPA                     = (Revenue x gross margin %) / Leads needed
Break-even ROAS             = 1 / gross margin %
Target ROAS                 = Break-even ROAS x 1.5
Required ad budget          = Leads needed x target CPA
```

These must exist before step 2: leads needed, max CPA, target CPA, required budget, target ROAS. Build three scenarios (pessimistic / base / optimistic) and plan against base.

Sanity-check the result against `references/benchmarks-global.md`: global Meta medians are CPA $18-38 and ROAS 1.93-2.79. Healthy ROAS rule of thumb is 3x at roughly 30% gross margin, 5x+ for commodity margins, and 1.5x can work for high-margin digital products. If your required CPA sits far below the bottom quartile for your industry and market tier, the plan is not achievable as written.

### 2. Regional cost baseline

Set the expected cost floor per market before allocating budget:

| Market tier | Examples | Meta CPM | Meta CPC |
|-------------|----------|---------:|---------:|
| Tier 1 premium | United States | $15-25 | $1.40-1.85 |
| Tier 1 | Canada, Australia, UK | $10-22 | $1.10-1.80 |
| Tier 1 Western EU | Germany and Western EU, Japan | $5-15 | $0.90-1.50 |
| Tier 2 SEA lead | Singapore | $6-12 | $0.70-1.20 |
| Tier 2 | Thailand, Malaysia, Indonesia, Philippines | $2-7 | $0.20-0.80 |
| Tier 2 LATAM | Brazil and Latin America | $2-6 | $0.30-0.70 |

Full table in `references/benchmarks-global.md`. Never blend Tier 1 and Tier 2 into one reported CPA — the average describes no real campaign.

### 3. Seasonality adjustment

| Period | Median CPM | vs average |
|--------|-----------:|-----------:|
| January | $15.74 | -21% |
| Q1 average | $18.29 | baseline |
| Q2-Q3 | $18-21 | rising |
| October | $22 | +10% |
| November (peak) | $25.22 | +27% |
| December | $22.04 | +11% |
| Q4 average | $22.98 | +26% above Q1 |

Planning implications: front-load acquisition into Q1 if cash flow allows; reserve at least 30% more budget for the same volume in Q4, or shift brand spend into Q3. In US election years, political spend displaces commercial inventory and adds roughly 20-40% to CPM in the last 30 days before the vote.

### 4. Channel allocation

| Channel | % Budget | Amount | Objective | Primary KPI |
|---------|---------|--------|-----------|-------------|
| Meta | [%] | [amount] | [Lead/Conversion] | CPA, CTR, ROAS |
| Google Search | [%] | [amount] | High intent capture | CPA, conversion rate |
| TikTok | [%] | [amount] | Video-first prospecting | CPA, view rate |
| YouTube | [%] | [amount] | Awareness, consideration | CPM, view rate |
| Pinterest | [%] | [amount] | Discovery for home/fashion/DIY | CPA, saves |
| LinkedIn | [%] | [amount] | B2B targeting | CPL, MQL rate |
| **Total** | **100%** | | | |

Starting archetypes, to be adjusted by real data:

- **DTC ecommerce:** Meta 45% + Google (Search plus Shopping) 30% + TikTok 25%
- **Lead gen B2C services:** Meta 50% + Google Search 30% + TikTok 20%
- **B2B SaaS:** Google Search 45% + LinkedIn 30% + Meta 15% + YouTube 10%
- **Creator, course, coaching:** Meta 50% + YouTube 25% + TikTok 25%
- **Home, decor, fashion:** Meta 45% + Pinterest 25% + Google 30%

Rule: a channel with proven performance takes 60-70%; a new channel gets a 15-20% test allocation. Do not spread evenly across channels you have never run.

**LinkedIn changes the math for B2B.** At $30-100+ CPM versus Meta's $13-20, LinkedIn costs roughly 2-5x more per impression. It is often still correct for B2B because the targeting precision and deal size justify it — but B2B/SaaS CPL benchmarks are $80-300, so validate against LTV in `10-reverse-kpi-global` before committing more than a test budget.

### 5. Funnel allocation

| Stage | % Budget | Audience | Creative | Primary KPI |
|-------|---------|----------|----------|------