Co-marketing for startups: borrow an audience, close it in a DM

Co-marketing for startups is two small companies promoting each other to their own audiences on the same day, splitting the production work and keeping their own customers. At indie scale there is no channel program and no revenue share. There is a direct message, a shared doc, five agreed fields, and two publish buttons pressed the same morning.

The advice written about this assumes otherwise. Search the term and you get glossary entries defining co-branding, agency posts about partner enablement, and enterprise programs built around partner portals, market development funds, and co-selling motions with a cloud provider. All of that is real. None of it survives contact with a founder who has 400 subscribers, no partner manager, and two spare afternoons.

What follows is the version that works at that size: the test for whether a partner is worth approaching, the four plays that actually close, a one-pager short enough to paste into a message, the rule for keeping it fair when one side is visibly bigger, and how to tell afterwards whether it did anything.

Why peer partnerships beat influencer outreach

For a B2B product with no budget, a peer partner outperforms an influencer because you are borrowing intent rather than attention. Someone who already pays for a tool in your category has proven three things a follower has not: they have the problem, they buy software to solve it, and they trust the person recommending you enough to have handed them a credit card.

The economics differ too. An influencer collaboration costs money or a great deal of relationship capital, and it ends when the post falls off the feed. A co-marketing swap costs an afternoon of production and leaves behind assets that keep working: a docs page, an integration listing, a joint article, a partners page entry. If you are weighing the two, the honest comparison is in micro-influencer collabs for small startups, which is the right play when your buyer is a consumer rather than an operator.

There is a 2026 argument on top of the traffic one. Answer engines have become a real discovery surface, and they do not cite the same sources: the analysis in 5W's State of AI Citations 2026 found Wikipedia and Reddit together supplying more than a quarter of US ChatGPT citations, while Perplexity leans on attributed pages and draws on roughly three times as many sources per answer as ChatGPT. A joint benchmark or teardown published under two named founders on two domains is exactly the shape of thing those systems reach for. One paid shoutout is not.

How do you find a co-marketing partner?

Start from your users' existing stack, not from a category list. Write down the five tools your current users already pay for that are roughly your size, and find the founder behind each one. Then run each candidate through four checks. A partner has to pass all four, and most fail on the third.

The check that saves the most wasted weeks is the third one. Founders quote follower counts and subscriber totals because those are the numbers they know. The number that predicts whether a swap moves anything is engagement on the last promotional send, so ask for it directly. A partner who will not share an open rate is telling you what the open rate is.

Where you meet these people matters less than it seems, as long as the venue shows you who is currently shipping. Launch communities, indie maker platforms, and founder directories all work, and the free channels that carry first-users traffic carry partnership conversations too: Product Hunt Upcoming, Indie Hackers, and the busier startup subreddits. Browsing the founder directory here does the same job with less noise, because every profile is attached to a live product and a public record of helping other people launch.

The four indie co-marketing plays

Almost every startup co-marketing deal that closes is one of four shapes. Pick by what you have, not by what sounds most ambitious.

If both audiences are small, joint content is the only play with a floor. A bundle promo sent to two lists of 200 produces a handful of clicks and a slightly embarrassed follow-up message. A joint benchmark published on both domains keeps earning after launch day, which is the same logic behind newsletter swaps for founders: the swap is the distribution, but the artifact is the asset.

One caution on the email half of any of these. If a partner sends to their list on your behalf, the FTC's CAN-SPAM compliance guide is explicit that both the company whose product is promoted and the company that actually sends the message can be held legally responsible, and that opt-out requests must be honored within 10 business days. You cannot hand that obligation to a partner along with the copy.

The partnership one-pager that fits in a DM

The reason most indie partnerships stall is not rejection. It is that both founders agreed enthusiastically to something neither of them wrote down, and three weeks later nobody is sure who was producing the email. Five fields fix that, and all five fit in a single message.

Send it with your half already produced. The gap in reply rate between "would you be interested in exploring a partnership" and a message containing the finished 120-word blurb, two screenshots and a proposed date is not subtle, because the second one moves the decision from whether to collaborate down to whether to spend twenty minutes. Open with one specific observation about their product that proves you have used it, then the five fields, then stop writing.

Keep the whole thing in one shared document you can both edit. No deck, no formal agreement, no scheduled kickoff call. A one-off promotional swap with no money moving does not need a contract. It needs a written yes and a list of who is producing what.

Keeping the split fair when the sides are uneven

You will almost always be partnering with someone bigger or smaller than you, and pretending otherwise is what makes the smaller side feel used and the larger side feel taken advantage of. Name the asymmetry in the first message and price it in work.

The rule that has held up for me: the smaller audience contributes more production. If their list is four times the size of yours, you write the joint post, you build the landing page, you cut the screenshots, and they send an email and publish a link. That is a fair trade of the two things each side actually has, and it turns a conversation about status into a conversation about tasks.

Three failure modes worth naming before they happen:

• The vanishing second half. You publish, they get busy, your side of the swap never runs. Prevent it by publishing on the same morning rather than agreeing to a vague sequence, and by making the smaller commitment the one that goes first.
• Mismatched offers. One side gives 40 percent off, the other offers the free trial everyone already gets. Compare the two offers the way a reader would before either goes out.
• The undisclosed sweetener.If money, credits, free accounts, or equity changed hands to make the promotion happen, that connection has to be disclosed. The FTC's Endorsement Guides treat an undisclosed material connection as deceptive, and one plain sentence costs nothing.

The reciprocity underneath all of this is the mechanic Favors.dev formalizes. It is a founder marketing co-op with a points economy: you earn points by doing verified marketing favors for other founders, and you spend them requesting help back. The reason that works is the reason a good co-marketing deal works. A stranger who has already helped you once is a far easier partner than a stranger who has not, and a public record of who actually delivers removes most of the guesswork from the fourth check above.

Did the partnership actually pay?

Judge a co-marketing swap on three things, in this order: qualified signups attributable to the partner link, assets that still exist a month later, and whether the partner would run it again. Raw traffic is the least useful number you will be tempted to look at.

The mechanics take about ten minutes. Give each side a distinct tagged link, agree in advance on the one number you will both report, and put the comparison date in the one-pager so the review actually happens. Then judge it against a deliberately low bar, because the first partnership is a test of the relationship as much as of the channel.

• Signups from the partner link. The only number that settles the argument. Compare it to what the same afternoon of work would have produced spent on anything else.
• Activation, not just registration. Borrowed audiences often sign up at a good rate and activate at a poor one. If they never reach first value, the shared-audience test was wrong, not the execution.
• The residue. A docs page, an integration listing, a partners page entry, an indexable joint article. These keep compounding and are the real return on a low-traffic swap.
• The second yes. Ask directly whether they would run another one. A partner who says yes without hesitating is worth more than the campaign was.

One partnership is a data point, not a channel. Three that all produce nothing means the shared-audience test is being applied too loosely, which is a fixable problem. Where partnerships sit against everything else competing for the same afternoon is laid out in the GTM strategy guide for solo founders.

Frequently asked questions