Load: ../plan-campaign/references/unit-economics.md, ../plan-campaign/references/company-types.md (the brand's type).
Inputs: the goal (metric, number, date) as the marketer stated it; budget; AOV / ARPA / first-job value; margin or contribution inputs; repeat behaviour if shown; baseline for the same period (marketer or past reports); conversion rates from the brand's history (learnings, past performance reads) or labelled base rates; stock or capacity.
Do — part 1 (step 2):
- Fill the economics box. Every line gets a source: own history · marketer · base rate (rule of thumb) · assumption. Unknown margin → a labelled assumption with "confirm margin" first in What we couldn't check (asked only when the plan turns on it).
- Baseline → incremental target (never credit the campaign for the baseline); implied CAC = budget ÷ incremental target vs history or labelled base rates; stock or capacity vs the goal. Pre-revenue: no allowable CAC — record the marketer's provisional cap. Do — part 2 (step 5, after the channels and split exist):
- Break the incremental target into per-leg targets with the chain that produces each (list × conversion; spend ÷ CPA; codes from creators), each with a "readable by" date.
- Run conservative / expected / optimistic using the low / middle / high end of each assumption.
- Check the forecast against stock (days of cover, inbound) or capacity (open slots).
- If the expected scenario misses: write the renegotiation case (gap in numbers; what more budget, more time, a smaller goal or a different goal metric each buys; one recommendation). Return it to the playbook to show the marketer before planning further.
Output: the plan's ### Goal chain, ### Economics, ### If the goal needs changing sections, and the assumptions list.
Never: invent the goal, a margin or a conversion rate and present it as the brand's; pick the renegotiation lever for the marketer.