agentleFS
minhnv0807minhnv0807/fullstack-mkt-skills/skills/en/61-budget-planning-global/SKILL.md

61-budget-planning-global

Skill644 starsChanged 6 months agoMITView on GitHub

Use when a marketing BUDGET has to be planned or managed in USD — allocation by bucket, channel, and month, test versus scale split, scale-up and stop-loss thresholds, reserve, a plan-versus-actual tracker, and a review cadence. Trigger on 'budget planning', 'marketing budget', 'split the budget across channels', 'stop loss threshold', 'quarterly budget', 'we keep overspending and nobody notices'. Also use when a number has been approved and nobody has divided it up. Not for — the revenue-backward math that produces the number, see `10-reverse-kpi-global`; paid media allocation only, see `54-media-plan-global`; paying an external vendor, see `67-agency-vendor-brief-global`.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Read .agents/product-marketing-context-global.md, the marketing plan, and the previous period's report if available.

    line 34
  2. Distinctive:

    Budget follows performance: Do not stay rigid to the opening plan.

    line 43
  3. Distinctive:

    Always hold 10-15% in reserve: Never allocate everything on day one.

    line 44

≈ “Common” is an estimate: at least 1 in 5 of the 372,429 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Budget Planning (Global)

A budget is not a fixed number — it is a decision tool: where you place bets and when you pull back. The leader must know total spend, ROI per channel, what to increase, and what to cut. Run 10-reverse-kpi-global first to derive the minimum budget from the revenue target before allocating anything.

Information gathering

Read .agents/product-marketing-context-global.md, the marketing plan, and the previous period's report if available. If missing, ask up to 4 questions:

  1. Total budget and planning period? How much, for a month, a quarter, or a year (USD)?
  2. Revenue target for the period? Needed to compute marketing as a percent of revenue and the minimum viable budget.
  3. Active channels and last period's ROI? Meta / TikTok / Google / email + SMS / affiliates / agency — which has the best ROAS or CPA?
  4. Any large campaign needing a reserve? Launch, Black Friday / Cyber Monday, or another seasonal peak.

Principles

  1. Budget follows performance. Do not stay rigid to the opening plan. Channels that work get more; channels that do not get cut.
  2. Always hold 10-15% in reserve. Never allocate everything on day one. The reserve covers unexpected opportunities and incidents.
  3. Every dollar has a measurable KPI. "Branding you cannot measure" is not an acceptable line item.
  4. Do not spread evenly. Concentrate on the best-performing channel. New channels get a minimum test allocation (10-15%) first.
  5. Plan backwards from revenue, not from channels. The budget is the output of a revenue equation, not an arbitrary number.
  6. Do not divide the annual budget evenly across quarters. Median Meta CPM in Q4 runs roughly 26% above Q1 (references/benchmarks-global.md), so identical monthly spend buys materially less reach in Q4. Either front-load acquisition into cheap Q1 months or explicitly reserve a larger Q4 allocation for the same volume.

Workflow

1. Choose the budget-setting method

MethodHow it worksStrengthWeakness
Percent of revenueBudget = X% of expected revenueSimpleIgnores growth stage
Backwards from targets (recommended)Set revenue > derive CAC/CPL > derive spendLogical, evidence-basedNeeds historical CAC/CPL
Competitive benchmarkMatch category spend normsAvoids under-investmentIgnores your own conditions

Seven-step backwards calculation (full detail in 10-reverse-kpi-global):

1. Target revenue
2. / AOV                        = orders needed
3. / close rate (30-50%)        = hot leads needed
4. / nurture CVR (50-60%)       = warm leads needed
5. / lead capture CVR (20-40%)  = total reach needed
6. total leads x CPL            = minimum ad budget
7. + content + tools + agency   = total marketing budget

Build 3 scenarios: Conservative (CVR below category median, CPL above), Base case (at median), Optimistic (post-optimization). Plan against Base; prepare cashflow against Conservative.

2. Allocate by bucket

BucketSuggested sharePurpose
Paid ads (performance)50-65%Meta, TikTok, Google, retargeting
Content production10-15%Video, design, copywriting
Creators / UGC / affiliates5-10%Social proof, third-party reach
Agency / freelancersPer scopeIf used
Tools / software2-5%Ad tools, design, email/CRM, analytics
Campaign reserve10-15%Opportunities and incidents — NOT pre-allocated

3. Allocate by channel

Starting split (adjust to your category and your own data):

ChannelShare of ad budgetObjectiveExpected CPA / ROAS
Meta Ads (Facebook + Instagram)45-60%Lead / purchase
TikTok Ads15-20%Awareness / lead
Google Search + Shopping15-25%High-intent capture
Email + SMS (Klaviyo, Mailchimp, Attentive)5-10%Retention, nurture, retargeting
New channel (test)10-15%Validate

For B2B, shift toward LinkedIn and Search — LinkedIn CPM runs $30-100+, so plan a higher CPL and a longer payback window. Allocation logic: channels with strong ROAS or CPA get a larger share; new channels get the minimum test allocation first; never divide evenly.

4. Split by funnel and by test vs scale

Standard performance split inside each channel:

PurposeShareNote
Testing (new creative + audiences)30%Always keep testing — creative fatigue arrives fast
Scale (confirmed winners)50%Only scale campaigns with winning data
Retargeting15%Warm audience — cheapest CPA in the funnel
Lookalike / similar audiences5%Expansion from converter seeds

By funnel: TOFU (reach/awareness), MOFU (lead/nurture), BOFU (conversion/retargeting). The ratio depends on stage: a launch skews TOFU, a mature program skews BOFU and retention.

5. Allocate by campaign and month

CampaignMonthBudget (USD)ObjectiveKPI
Always-on (retargeting + nurture)Full period
Campaign A (launch or seasonal)
Testing budgetFull periodNew channels or angles
Reserve—10-15%

Seasonality overlay before locking monthly numbers:

PeriodMedia cost vs Q1Planning implication
January~-21% below Q1 averageCheapest acquisition window — stock budget here if cashflow allows
Q2-Q3RisingNormal operating cadence, build audiences for Q4
October~+10%Start of the expensive window
November~+27% (peak)Reserve extra budget or accept lower volume
Q4 average~+26% above Q1Do not plan Q4 at Q1 unit costs

In US election years, political spend displaces commercial inventory and adds roughly 20-40% to CPM in the last 30 days before the vote — build that into Q4 plans for US-targeted campaigns.

6. Scale-up and stop-loss rules

Increase budget when (BOTH conditions hold):

  • ROAS above 1.5x target for 7+ consecutive days, OR CPA at or below 70% of target with sufficient volume
  • Audience not yet saturated (frequency below 2.5)
  • Increment: +20-30% per step, never a sudden doubling

Cut or pause when (stop-loss):

ConditionAction
ROAS below break-even for 14 days after enough data (break-even on contribution margin before ads, not gross margin)Cut the channel or campaign, move budget to what works
CPA above 1.5x target for 7 daysChange creative or audience
CPA above 3.0x target once spend >= 3x target CPA (below that, not enough data; house heuristic ladder in 03-performance-eval-global)Propose a pause (rule 7 draft block, named approver); replace creative or audience before restarting
Frequency above 4 (urgent tier; heuristic)Audience saturation — stop scaling, refresh creative or expand audience; frequency alone never pauses an ad set
Spend above 110% of the monthly planCap daily budgets, review the allocation

Weekly CPA review rule (expressed relative to your own target, since absolute CPA varies 3-5x between regions and industries):

CPA vs targetFrequencyVerdict
<= 1.0x target<= 2.5WIN — scale +20-30%
1.0-1.5x target<= 2.5Watch — monitor
> 1.5x target, under 7 daysanyWatch — keep monitoring, no verdict yet
> 1.5x target for 7 daysanyOptimize — change creative or audience
> 3.0x target (spend >= 3x target CPA)anyPropose a pause through rule 7 (draft block, named approver)

Step 0: spend below 3x target CPA = not enough data, no verdict. Source: house heuristic ladder in 03-performance-eval-global (heuristic).

Set the target itself using references/benchmarks-global.md for your region and industry, not a global average. Never increase budget while CPA is deteriorating — fix creative or tracking first.

7. Tracker and review cadence

MonthPlanned budgetActual spendVarianceROAS / ROINote
Month 1
Month 2
Month 3
Quarter total
CadenceWorkOutput
WeeklyReview CPA and ROAS per channel, apply the step-6 decision rulesImmediate adjustment when a threshold trips
MonthlyReview the full allocation, reconcile plan vs actualRebalance for next month
QuarterlyReview overall allocation strategy and LTV:CACStrategy adjustment

Unit economics health — LTV:CAC:

RatioMeaningAction
< 1:1Losing money on every customerStop scaling, fix product or price
1:1 to 3:1Break-even to viableOptimize conversion, raise AOV
> 3:1HealthyBegin scaling
> 5:1StrongIncrease ad budget aggressively

Output structure

File name: budget-plan-[brand-name]-[YYYYMMDD].md — contains: I. Budget overview (total, revenue target, marketing as % of revenue, period) · II. Bucket allocation · III. Channel allocation · IV. Test vs scale split · V. Campaign and monthly allocation with seasonality overlay · VI. Decision rules and stop-loss thresholds · VII. Plan-vs-actual tracker · VIII. Review cadence.

Related skills

  • 10-reverse-kpi-global: derive the minimum budget from the revenue target — run before allocating.
  • 00-marketing-plan-global: the budget plan is the detailed version of the plan's budget section.
  • 54-media-plan-global: turn the ad budget into a concrete campaign and ad-set level media plan.
  • 21-ads-audit-global: audit a channel that repeatedly trips the stop-loss thresholds.
  • 07-marketing-report-global: budget reconciliation inside the monthly report.

Quality checklist

  • Budget derived backwards from the revenue target, with 3 scenarios
  • All 3 allocation layers present: bucket > channel > campaign/month
  • Test vs scale split explicit (testing 30 / scale 50 / retargeting 15 / lookalike 5)
  • 10-15% reserve held back, not pre-allocated
  • Scale-up and stop-loss thresholds use concrete numbers (ROAS, CPA vs target, frequency, days)
  • New channels receive only a 10-15% test allocation
  • Every budget line has a measurable KPI attached
  • Monthly plan accounts for Q4 media inflation rather than dividing the year evenly
  • Plan-vs-actual tracker exists by month
  • Weekly, monthly, and quarterly review cadence with defined outputs
  • LTV:CAC checked before any scale decision

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.