Lesson 83.3 gave you the five-part sponsor filter and one hard rule: one failure is a decline. This appendix turns that lesson into a one-page policy a future operator can follow under revenue pressure — selection, member benefit, disclosure, inventory caps, and decline criteria, plus a filled niche-tool example.

The blank policy

# COMMUNITY-PARTNER-POLICY.md — [Community/publication], [date]

## Qualifying criteria (must pass ALL five)
- Relevance (members discuss it unprompted?): ___
- Tested utility (who tested on real work, when?): ___
- Transparency (disclosable in one plain sentence?): ___
- Separation (sponsor veto? never / flaw-note rule?): ___
- No-go check (clear of banned categories?): ___

## No-go list (named categories, no euphemisms)
- ___

## Member benefit (what members receive — required)
- ___

## Inventory cap (scarcity on purpose)
- Max sponsor slots per month: ___ / max affiliate mentions per issue: ___

## Disclosure language (copy-paste exact wording)
- Sponsored segment: "___"
- Affiliate link: "___"

## Editorial boundary
- Who writes/reviews: ___ / flaw-note rule: ___

## Approval owner + review cadence
- Owner: ___ / cadence: ___ / active-partner log kept at: ___

## Decline criteria + script
- Auto-decline when: ___ / script: "___"

Three fields do the heavy lifting. Member benefit is the deal's reason to exist — if you cannot name what members receive in one specific sentence, there is no deal, only a rental. Inventory cap keeps attention scarce: one slot a month stays special; weekly slots become wallpaper members learn to skip. Exact disclosure wording removes improvisation at the worst moment — under deadline, with a check waiting, nobody writes brave sentences. Write them now.

Filled example: niche filing-data tool (illustrative fictional example — subscriber and commission numbers are teaching placeholders)

A weekly research memo with 2,100 subscribers filled the policy like this:

# COMMUNITY-PARTNER-POLICY.md — The Filing Memo, 2026-10-04

## Qualifying criteria
- Relevance: tool serves solo analysts building citable briefs (our readers)
- Tested utility: two members ran it on live filings for 3 weeks; notes on file
- Transparency: one-sentence disclosure up top, commission % stated
- Separation: editor writes teardown from draft before sponsor sees it;
  sponsor briefed after, no veto, flaws published
- No-go check: clear — tested, evaluable, no data harvesting

## No-go list
- Untested financial products / data-harvesting browser tools /
  anything members cannot independently evaluate / "partner" language
  without naming payment

## Member benefit
- One worked teardown/month using the tool on a live filing + 20% member
  discount + archived Q&A; dues stay $12 because this slot funds the library

## Inventory cap
- Max 1 sponsor slot/month / max 2 affiliate mentions per issue

## Disclosure language
- Sponsored: "Paid partner — they funded this teardown; conclusions are
  ours; flaws noted below."
- Affiliate: "Affiliate link — we earn 20% if you subscribe; we recommend
  it because we use it; re-tested September 2026."

## Editorial boundary
- Editor writes and reviews; reviewer free to note flaws; sponsor never
  previews conclusions

## Approval owner + cadence
- Owner: editor / quarterly re-review of every active partner /
  public log: partners page with product, cut, last re-test date

## Decline criteria + script
- Auto-decline: fails any one of the five; exceeds inventory cap;
  asks for conclusion approval or vague disclosure
- Script: "Thanks — this doesn't pass our tested-utility bar for members,
  so we're passing. Happy to reconsider after two members run it for a month."

Notice the economics line inside member benefit: the slot funds the source library, which is why dues stay at $12. Members can see the exchange. The decline script is pre-written because, as Lesson 83.3 warned, exceptions are the policy — write the no before the check arrives.

Rules and failure modes

• Keep the public affiliate log current: product, commission rate, last re-test date. A log you will not publish is a deal you should not sign.
• Re-test every partner on cadence. Tools decay, pricing changes, ownership turns over — last year's honest recommendation is this year's unexamined ad.
• The "just this once" exception for a big check is the failure mode. Cap breaches compound: the second weekly slot is easier to sell than the first, and the room learns the schedule before you admit it changed.
• For tier and disclosure patterns worth borrowing while drafting, browse the Patreon Creator Hub; for how packaging and fee pressure shape partner math, compare the Paddle pricing strategy guide and Gumroad pricing.