Lesson 78.4 gave you a direct relationship you own — an email list that returns without asking a feed for permission. Partnerships are the next borrowed route worth earning: a credible introduction to someone else's audience, where trust transfers because the joint offer is genuinely useful. This lesson teaches you to choose partners by shared value, not follower counts.

The vocabulary, in plain language

• Partnership: a deliberate agreement between two independent projects to create value neither could create alone — shared audience, shared asset, shared follow-up.
• Partner audience: the specific group the partner can actually reach and influence — readers who open, customers who buy, members who show up. Not their total follower number.
• Co-marketing: a joint message to overlapping audiences — a shared guide, webinar, sample, or announcement where both sides contribute and both sides benefit.
• Integration partner: a product whose tool connects to yours, so the combined workflow is better than either alone. Think a research template that lives inside a note-taking app.
• Referral partner: a trusted party who recommends you at the right moment and is recognized for it — a consultant, shop owner, teacher, or newsletter writer whose word carries weight.
• Audience overlap: the specific people who plausibly need both offers. "Small research teams who write source-backed briefs" is an overlap. "People interested in AI" is not.
• Mutual value: each side gains something real — qualified users, useful content, better retention — proportional to what each side contributes.

The mental model: a partnership is a bridge between two desks. People cross only if the other side holds something they already want, the bridge is clearly signposted, and the landing desk is ready to serve them.

The partnership test: five checks before you pitch

Most partnership failures share one cause: one side wanted access, the other side got work. Run these five checks honestly before you write any outreach.

1. Each side brings a real asset. An audience that opens, a product that works, a template people save, a room people attend. If either side brings only "exposure" or "potential," there is no partnership yet — there is a request.

2. The overlap is specific. Name the shared person, moment, and job. "Solo analysts who must cite filings every Friday" lets you design a joint offer. "Knowledge workers" lets you design nothing. Write the overlap in one sentence and ask both sides: would you recognize this person if they replied?

3. The joint offer is useful on its own. A co-marketed checklist, a sample brief built on the partner's question, an integration that removes three manual steps. If the joint thing would embarrass you without either logo, the logos will not save it.

4. Roles are clear. Who provides the audience, who provides the asset, who writes, who sends, who answers replies, who owns the destination? Put names beside each. Vague "let us collaborate" produces vague follow-through.

5. The destination can serve the newcomer. A partner who sends forty qualified readers to a broken signup, a generic homepage, or a waitlist with no next step will not send a second batch. Lesson 72.2's message match applies here: the landing page must continue the joint promise with proof, setup, and one action.

If any check fails, redesign or decline. A polite no protects both reputations better than a thin joint webinar nobody replays.

Desk plus newsletter: a worked example

Take the Research Desk — a source-linked research product for analysts who must produce cited briefs — and a niche newsletter read by 2,400 independent researchers who value evidence over hype.

The newsletter's asset is a partner audience that already trusts its curation. The Desk's asset is a genuinely useful joint object: a one-page "footnote-check template" plus a worked sample brief on a company the readers asked about. The overlap is specific: solo researchers who cite filings weekly and dread unsourced claims.

The co-marketing shape is small: the newsletter publishes the sample with method notes; the Desk hosts the blank template plus a three-minute setup video. Roles are explicit — the writer edits for accuracy, the Desk founder answers replies for one week, the newsletter sends once with a tracked link to the template page, not the homepage. The destination serves: template downloads in one click, sample beside blank, email follow-up only with permission (Lesson 78.1).

The referral variant is equally tight. The newsletter becomes a referral partner only after readers actually use the template — a tracked link that pays a flat thank-you per activated workspace, disclosed plainly. No pop-up, no forced share, no "invite ten friends for a badge." If the template does not earn a forward on merit, the referral layer never launches.

What this example rejects matters as much as what it accepts: no logo-swap, no audience-size bragging, no joint "AI trends" panel with no object to take home. Small, specific, servable.

Exercise: build the shortlist before the pitch

Create PARTNER-SHORTLIST.md. Five candidates, no more, each judged by the test above.

# PARTNER-SHORTLIST.md — [Project]

## Candidate 1 — [name / newsletter / product / community]
- Partner audience + asset: ___ (size that opens/uses, not vanity total)
- Overlap (one specific sentence): ___
- Joint offer (useful object): ___
- Roles: they ___ / we ___
- Destination + servability check: ___ (page, proof, one action)
- Contact rationale (why now, why us): ___
- Reason NOT to pursue (or what must change first): ___

## Candidate 2 — (same fields)
## Candidate 3 — (same fields)
## Candidate 4 — (same fields)
## Candidate 5 — (same fields)

## Hypothesis to test first
- One partner + one joint offer + send date: ___

Worked mini-example — creator-led niche publication: candidates include a citation-tool maker (integration partner: export-to-brief button), two newsletters with overlapping researcher readers (co-marketing: shared teardown), a freelance analyst collective (referral partner: client-ready template), and a local business network (deferred — overlap too vague until the offer narrows). Each carries its no-pursue reason; the first test is the citation-tool integration because the joint object already exists.

Finish line: a PARTNER-SHORTLIST.md with five candidates, each with audience, overlap, mutual offer, roles, destination, contact rationale, and one explicit reason not to pursue or what must change first.

Verify quickly: cover the partner names and read only the overlap lines. If a stranger cannot tell which partner matches which overlap, the overlaps are too generic — rewrite until each fits exactly one partner.

Common failure mode: the follower chase — pitching the biggest audience regardless of overlap, then designing the joint offer around the pitch. Overlap first, offer second, outreach last. Reversed order produces polite calls and zero crossings.

Check your understanding

1. Define partner audience, co-marketing, integration partner, referral partner, and audience overlap in your own words. 2. Why does each side need a real asset before the partnership test can pass? 3. What breaks when the destination cannot serve the newcomer's next step?

Next

One trusted bridge is useful; a place where strangers already search is different. Lesson 79.2 moves to marketplaces and ecosystems — where intent already exists, but the format, quality, support, and review bar belong to someone else.