Media Planning and Budget Allocation SKILL.md

This is the exact file agents read, including the credit header and licence notice. Save it as media-planning-budgeting/SKILL.md.

---
name: media-planning-budgeting
description: "Build media plans and budget allocations from unit economics: maximum new-customer CAC from margin, lifetime value and payback, channel baselines labelled by evidence, response curves and diminishing returns, channel roles across search, shopping, social, video, retail media and affiliates, scenarios, pacing, reallocation rules and a ring-fenced test budget. Use whenever someone asks how much to spend on ads, how to split budget across channels, a media plan, marketing budget allocation, CAC targets, payback period, or how to scale paid media profitably."
license: CC-BY-4.0
metadata:
  author: Afeez Adebayo
  author_url: https://afeezadebayo.com
  author_linkedin: https://www.linkedin.com/in/afeez-adebayo/
  author_x: https://x.com/za_icon
  source: https://afeezadebayo.com/skills-md/marketing/performance-marketing/media-planning-budgeting/
  category: marketing
  subcategory: performance-marketing
  version: "1.0.0"
  created: "2026-10-10"
  updated: "2026-10-10"
---

# Media Planning and Budget Allocation

Most media plans are last year's split with a percentage added, or whatever each platform's dashboard says is working best. Both approaches reward channels that are good at claiming credit rather than channels that add customers. A sound plan starts from business goals and unit economics, sets how much the business can pay for a customer, allocates across channels using the best evidence available and keeps budget aside to test.

This skill builds media plans and budget allocations for paid search, shopping, paid social, video, retail media and affiliates, with scenarios, pacing and a test budget.

## Operating principles
1. **Start with economics.** Gross margin, lifetime value and payback period set the maximum cost per customer.
2. **Plan for incremental customers,** not platform conversions.
3. **Respect diminishing returns.** The next pound in a channel usually returns less than the last one.
4. **Fund the full funnel deliberately.** Demand creation and demand capture work together; cutting one often hurts the other later.
5. **Ring-fence testing.** Keep 10 to 20% of budget for experiments so the plan improves each quarter.

## Unit economics first
| Input | Example | Use |
|---|---|---|
| Average order value or contract value | £85 | Revenue per conversion |
| Gross margin | 55% | Profit before marketing |
| Repeat purchases in 12 months | 1.6 orders | Lifetime value |
| 12-month gross profit per customer | £85 × 1.6 × 55% ≈ £75 | Ceiling for acquisition cost |
| Target payback | 6 months | How much of that value you can spend up front |
| Maximum new-customer CAC | Depends on payback and cash | Sets bids and budgets |

## When to use, and when not to
Use it for annual and quarterly planning, launching in a new market, adding a channel, when growth stalls, and when finance asks how much to spend.

Don't use it to run a specific channel (use the channel skills) or to measure incrementality in depth (`incrementality-testing`, `marketing-mix-modelling`).

## Inputs to ask for, in one message
1. Business goals for the period: revenue, new customers, profit, market share.
2. Unit economics: order or contract value, margins, repeat rate, LTV, payback target.
3. Spend and results by channel for 12 to 24 months, monthly.
4. Any incrementality evidence: lift tests, geo tests, MMM.
5. Seasonality and planned launches or promotions.
6. Constraints: cash, minimum spends, contracts, team capacity.

## Workflow

### Step 1. Set the target
Translate goals into the number of new customers or pipeline needed and the maximum CAC the economics allow.

### Step 2. Baseline each channel
For each channel: spend, platform results, your own data results (new customers, revenue), and any incremental evidence. Label the confidence of each estimate.

### Step 3. Estimate response curves
Use MMM outputs if available; otherwise use past spend changes and tests to estimate how results change with spend. Expect flattening at higher spend.

### Step 4. Allocate
Fund proven, efficient channels up to the point where marginal CAC hits the limit; add demand creation channels with test-based evidence; ring-fence a test budget.

### Step 5. Build scenarios
Conservative, expected and stretch plans, each with spend, expected new customers, blended CAC and risks.

### Step 6. Plan pacing
Monthly or weekly budgets following seasonality and promotions; rules for moving budget between channels based on leading indicators.

### Step 7. Define the test plan
Which questions the test budget answers this quarter (new channel, retargeting incrementality, brand search value).

### Step 8. Agree review cadence
Weekly pacing checks, monthly allocation reviews, quarterly re-plans with new test results.

## Channel roles
| Channel | Typical role | Watch out for |
|---|---|---|
| Brand search | Capture existing demand | Paying for customers who would come anyway |
| Non-brand search | Capture active intent | Rising costs at scale; broad match drift |
| Shopping and Performance Max | Capture product demand | Mixing existing customers and brand |
| Meta and TikTok prospecting | Create demand, reach new audiences | Over-crediting view-through conversions |
| Retargeting | Convert warm audiences | Often low incrementality |
| YouTube and CTV | Build awareness and consideration | Hard to measure without lift tests or MMM |
| LinkedIn | Reach B2B roles and accounts | High costs; measure pipeline, not clicks |
| Retail media (Amazon, others) | Capture marketplace demand | Halo effects on other channels |
| Affiliates and partnerships | Pay for performance | Coupon sites taking credit for organic sales |

## Budget plan template
```
| Channel | Role | Q spend | Expected new customers | Marginal CAC | Evidence (label) | Test planned |
| Total | | | | Blended CAC | | Test budget % |
```

## Estimating marginal CAC without an MMM
1. List past periods where a channel's spend changed meaningfully (a step up or down of 20% or more) while other channels stayed steady.
2. Compare new customers (from your own data) before and after each change.
3. Marginal CAC ≈ change in spend ÷ change in new customers.
4. Repeat for several changes to see whether marginal CAC rises with spend.
5. Label the estimate "observational" and plan a test to confirm it.

| Channel | Spend change | New customers change | Marginal CAC | Confidence |
|---|---|---|---|---|
| Meta prospecting | +£20k/month | +210 | £95 | Medium (one clean period) |
| Non-brand search | +£15k/month | +90 | £167 | Low (seasonality overlap) |

## Reallocation rules
- Move budget weekly only on leading indicators you trust (new customers from your data, qualified leads), not daily platform ROAS.
- Cap any single shift at 20% of a channel's budget per fortnight to avoid resetting learning.
- If a channel's marginal CAC exceeds the limit for four weeks, reduce it and log the decision.
- If a test shows a channel is incremental, scale in steps and re-measure.

## Funnel balance check
| Share of spend | Demand creation (prospecting, video, CTV) | Demand capture (search, shopping, retargeting) |
|---|---|---|
| Warning sign | Under 20% for a growth-stage brand | Over 80% while brand search volume is flat or falling |
Flat brand search volume after cutting prospecting is an early sign that capture is eating into future demand.

## Planning calendar
- **Annual:** unit economics, targets, big tests, seasonal peaks.
- **Quarterly:** reallocation from test results and MMM updates.
- **Monthly:** pacing against plan, scenario tracking.
- **Weekly:** spend pacing and anomaly checks.

## Edge cases and judgement calls
| Situation | What to do |
|---|---|
| No incrementality evidence anywhere | Start with conservative allocations and run one geo or lift test per quarter |
| Cash-constrained startup | Shorter payback targets; favour capture channels; small structured tests for creation |
| Strong seasonality | Shift budget ahead of peaks to build audiences; plan for higher costs at peak |
| New market entry | Budget for learning; expect higher CAC early; set clear stop-loss rules |
| B2B with long cycles | Plan on pipeline and leading indicators; review quarterly, not weekly |

## Handoffs to related skills
- Current account health: `paid-media-audit`
- Incrementality evidence: `incrementality-testing`, `marketing-mix-modelling`
- Channel execution: `google-search-ads`, `performance-max-shopping`, `meta-ads`, `paid-social-tiktok-linkedin`
- Reporting against the plan: `paid-media-reporting`

## Worked example
A subscription coffee brand planned to grow new customers 30% with a fixed budget increase. Unit economics showed 12-month gross profit per customer of about £70 and a target payback of five months, setting a maximum new-customer CAC. Baselines showed brand search and retargeting claimed many conversions but a past geo test showed retargeting added little. The plan cut retargeting, moved budget to Meta and YouTube prospecting with a geo test to validate it, held brand search at a defensive level, and kept 15% for tests. The expected scenario hit the new-customer target at a blended CAC within the limit, with a stop-loss rule if prospecting CAC exceeded it for four weeks.

## Output format
```
# Media plan: <business>, <period>
## Goals and unit economics (max CAC, payback)
## Channel baselines with evidence labels
## Allocation and scenarios (conservative, expected, stretch)
## Pacing by month
## Test plan and budget
## Review cadence and reallocation rules
```

## Common mistakes
- Allocating by platform ROAS.
- Ignoring diminishing returns when scaling.
- No test budget, so the plan never improves.
- Cutting demand creation because it's hard to measure.
- Plans without margin or payback limits.

## Questions to ask before you plan
- What can the business afford to pay for a new customer?
- Which channels have incrementality evidence?
- What cash and capacity constraints apply?

## Self-check before you answer
- Is the plan built from unit economics and payback?
- Are evidence labels shown for each channel estimate?
- Are scenarios, pacing and a test budget included?
- Are reallocation rules clear?

## Further reading
- Google Meridian (open-source marketing mix model): https://developers.google.com/meridian
- Meta Robyn (open-source MMM): https://facebookexperimental.github.io/Robyn/

---
Created by Afeez Adebayo, Marketing, SEO & Growth Expert
Website: https://afeezadebayo.com
LinkedIn: https://www.linkedin.com/in/afeez-adebayo/
X: https://x.com/za_icon
Source: https://afeezadebayo.com/skills-md/marketing/performance-marketing/media-planning-budgeting/

Licence: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). You may copy, share and adapt this skill, including for commercial use, as long as you credit Afeez Adebayo and link to the source URL above. Keep this notice in any copy or adapted version.