Almost every community partnership is proposed as a swap of audiences, and almost every one of those produces nothing. Two owners announce each other, both check the signups the next morning, and both quietly decide that partnerships do not work.
They do work. The swap is just the one format that cannot, because of what it asks the other side's members to do.
Why the shout-out swap fails
A shout-out asks people to take an action on behalf of a stranger, on the strength of somebody else's word.
It looks like the fairest possible deal — equal reach, equal effort, nobody owes anybody. But reach was never the thing you were short of. Trust does not transfer by announcement, and an endorsement of someone the audience has never seen is worth about what it costs to write, which is nothing.
What does transfer is a person turning up and being useful in public. That is a completely different transaction, and it is the only one worth organising your growth around.
The filter that works: same person, different moment
The standard advice is to find communities that are adjacent but not competing. It is the wrong filter and it rules out your best partners.
"Adjacent" usually means two rooms that are vaguely about the same subject, whose members share nothing specific enough to act on. Meanwhile the obvious win gets screened out for looking like competition: a community of people about to do the hard thing and a community of people who have just done it are ideal partners, because each is the other's before and after.
So stop asking whether they compete with you and start asking whether their members are your members at a different point in the same journey. If they are, everything you say will land, because it is about something those people are actually in the middle of — which is the version of "borrow someone else's audience" that survives contact with reality, and the one worth reaching for when you are still looking for your first hundred members.
What to actually propose
Four formats do almost all the real work, and each one fails in a specific, predictable way.
| Format | What it costs you | How it usually fails |
|---|---|---|
| A guest session in their space | An hour, and answering the awkward questions honestly | You pitch, and that is the only thing anyone remembers |
| A co-hosted event | Two calendars and a promotion plan | Both sides assume the other one is promoting it |
| A written piece for their members | A day, and material you would rather keep | You send something generic you had already published |
| A shared space or joint cohort | Months, and genuine operational work | Nobody agreed who moderates it or who keeps the members |
The first row is where almost everyone should start, and it is the one people skip because it feels too small to bother with.
Asymmetry is the normal case
Someone always has the bigger room, and insisting on an even trade limits you to partners exactly your own size, which is the smallest pool available.
Offer to do more than your share and stop keeping score. If they are larger, you are asking for credibility rather than reach, and credibility is expensive to lend — the correct response to that is to make the hour genuinely worth their members' time, not to negotiate a matching favour. The partner who is generous first is the one who gets asked back, and the second invitation is where this starts paying.
If you are the larger side, the useful reframe is that you are not doing charity. A guest who is deep in a problem your members are facing is programming you did not have to make, which is the same reason events work at all.
Be a member before you are a partner
Most partnership requests arrive from people who have never posted in the community they are writing to. That is why they are ignored, and it is entirely reasonable that they are.
Join. Read for a few weeks. Answer some questions where you actually know the answer, and do not mention what you run. It costs almost nothing and it changes the proposal from a stranger's request into a note from someone the owner has already watched being helpful — which is the whole thing you were trying to manufacture with a shout-out.
Then make the ask specific and one-sided. Name the session, name what their members get, and do not attach a reciprocal demand to the first message. A proposal that requires them to also do something for you is a negotiation; a proposal that just offers something useful is easy to say yes to, and the return favour arrives on its own far more often than you would expect.
Your members are not inventory
The one hard rule: never trade member lists, and refuse it when it is proposed to you.
Your members joined a community, not a distribution agreement. Handing their contact details to somebody else — even somebody good, even in exchange for the same — is a betrayal of the specific thing they trusted you with, and no partnership return is worth being the owner who did that. The moment members suspect their presence is an asset you trade, the honest posting stops, and the honest posting was the whole product.
Everything legitimate is available without it. Appear in their room, let them appear in yours, and let people move under their own power — which is also the difference between a partnership and the referral mechanics your own members run.
One collaboration is a wasted setup cost
The single most common way this fails is doing it exactly once.
The first appearance in an unfamiliar community converts badly, because nobody knows who you are yet and most people were not there. The third one converts well, for the same reason a returning speaker draws a bigger room than a new one. Nearly everybody quits after the first, counts eleven signups, and files partnerships under things that do not work.
So plan any partnership as at least three occasions across a few months, with the same person and the same room, before you judge it. If that sounds like too much for one relationship, you have discovered the real constraint: this channel rewards a handful of deep partnerships and punishes a long list of shallow ones.
How to tell whether it worked
Signups on the day are the least informative number available and the only one anybody looks at.
People who are in the middle of something rarely act immediately; they remember you and arrive weeks later, through a search or a second mention. So the honest measures are slower: whether members from that community are still turning up three months on, whether anyone writes "I saw you in X" in an intro, and whether the other owner asked you back.
The last of those is the real scoreboard. Being asked back means you were useful to their members, and being useful to their members is the entire mechanism — measure it alongside the metrics that mean something rather than in a spreadsheet of one-day spikes.
Three things not to do
Do not send the same message to forty communities. It is recognisable on sight, it is the reason owners have stopped reading these, and one specific note to the right person beats all forty.
Do not open with "let's collaborate". It puts the work of inventing the idea on the person you are asking a favour of, and they will not do it.
Do not lead with money. A revenue share offered before anyone has seen you be useful reads as a purchase of access, and it prices a relationship that has not happened yet — the commercial arrangements in monetising a community work later, on top of something real, not instead of it.
The bottom line
Partnerships are not a growth hack with a bad reputation. They are a slow channel that most people run in the one configuration guaranteed to fail.
Pick people whose members are yours at a different moment, join their room and be useful in it before you ask for anything, propose the smallest real thing rather than the most impressive one, and plan to come back twice more. Then judge it on whether you were invited again, not on what happened the next morning — and if you only ever manage two of these properly, that will still be worth more than the forty introductions you did not send.