Why we are building RelayWonder

RelayWonder is a SaaS partner network, and we are building it because affiliate software solves the second problem a founder has and ignores the first. The second problem is tracking: who clicked, who paid, who is owed what. Rewardful, FirstPromoter, Tolt and a dozen others handle that part well. The first problem is recruiting: who should I invite at all, and why would they say yes? Nobody sells a good answer to a company doing $5,000 to $50,000 a month in revenue. Agencies are priced for bigger budgets, marketplaces are priced by quote, and the tools assume the partners already exist. That gap is the company.

What affiliate software already does well

We want to be fair to the tools we are not trying to replace. Every affiliate platform we tried in the last two years does the same job and does it competently. A partner gets a link. The link drops a cookie. A Stripe webhook fires when the visitor pays. A commission row appears, and at the end of the month someone exports a CSV and pays people. The differences are in the edges: whether the cookie is first-party, whether renewals are tracked, how refunds are handled, whether the payout CSV matches what PayPal expects.

That part has become a commodity. Rewardful and FirstPromoter both start at $49 a month, both integrate with Stripe, both export a PayPal or Wise payout file, and both take 0% of the commission (as checked on 2026-09-18) [1][2]. A founder choosing between them is choosing on taste, not capability. We have no interest in winning that comparison on features alone, and we do not think anyone should pay us for a slightly different link-and-cookie pipeline.

• Links and cookies: solved. First-party cookies with a 30 to 90 day window are standard.
• Stripe attribution: solved for first payments; renewals and refunds vary by vendor.
• Payout files: solved. PayPal and Wise mass-payout CSV exports exist in most tools [4][5].
• Program pages and partner dashboards: solved, with varying degrees of polish.

So when a founder signs up for one of these tools, they get a working pipeline and an empty partner list. The tool has done its job. The founder has not started theirs.

The recruiting gap nobody prices for

Here is what actually happens next, because we have watched it and we have done it ourselves. The founder opens a spreadsheet. They search for "best tools for X" and copy the authors into column A. They search YouTube for "[competitor] review" and copy the channel names into column B. On a Sunday they write forty cold emails. Three people reply. One asks for a free account. Nobody posts a link. Six weeks later the program has two partners, both friends of the founder, and $340 in attributed revenue. The $49 a month looks like a bad deal, and the program is quietly abandoned.

The industry has two answers to this, and both are priced for someone else. The first is an agency. Retainers we have been quoted personally started around $3,000 a month, before any performance fee, and the agency keeps the relationships when you leave. The second is a managed marketplace. PartnerStack has a large B2B partner marketplace and handles payouts on the platform, but pricing is quote-based (as checked on 2026-09-26) [3], and it is built for companies with a partnerships hire, not a founder who also does support.

Look at the "who finds the partners" column. In three of the four rows, the answer is either "you" or "someone you are paying a lot". A SaaS doing $20,000 a month cannot afford the agency and does not qualify for the quote. They get the empty list.

Why a SaaS partner network instead of another tool

The recruiting problem is not a software problem in the usual sense. You cannot write code that makes a stranger post a link. What you can do is change the shape of the market so that the people worth inviting are already visible, already verified, and already paid through the same place. That is the difference between a tool and a network, and it is why we are building a SaaS partner network rather than a better affiliate dashboard.

In a tool, every brand has its own private list of partners. A newsletter writer who promotes six SaaS products has six logins, six dashboards, six payout thresholds and six tax form uploads. Nobody can see her track record except each brand, separately, for its own product. When a seventh brand wants to invite her, they are starting from zero.

In a network, she has one account. Her results across brands are hers. A brand considering her can see that she has sent paying customers to similar products, what her conversion rate looked like, and whether any of it was refunded. She carries one balance per currency across every program, so a $12 commission from one brand and a $38 commission from another add up to a $50 payout instead of sitting below two separate thresholds forever. Four partner types fit the same model: creators, affiliates, publishers and media, and communities.

That structure does something for the brand that no tool can do. The invitation goes to someone with a history, the acceptance takes under a minute because the account already exists, and the first payout is not blocked by a tax form because the form was uploaded months ago for a different brand. The founder still decides who to invite. The network makes the decision possible.

Tracking that follows the subscription, not the click

We said tracking is a commodity, which is true for the first payment. Where tools diverge is everything that happens after. A SaaS customer is worth a stream of invoices, and a partner who is paid only on the first one is being paid on a fraction of what they created. So RelayWonder attributes at the Stripe invoice level: the first payment and every renewal are matched to the partner, and the commission is calculated on each invoice amount [6].

A worked example. A customer signs up through a partner link on a $49 a month plan with a 20% commission for 12 months. Month one: invoice $49, commission $9.80. Month two: same. In month five the customer upgrades to $149. The invoice is $149, so the commission that month is $29.80, with no manual adjustment. In month eight the customer disputes a charge and Stripe reverses it. A negative row of minus $29.80 appears in the ledger against the same partner, and the balance adjusts. Nothing is edited or deleted; the ledger is append-only, which means the partner and the brand see the same history and can audit any month.

Two more details we considered non-negotiable. First, personal coupon codes attribute with no click at all. A podcast listener who types the code at checkout three weeks later is credited to the podcaster, even though no link was ever clicked. Second, the defaults are conservative so that a brand does not pay out money it will later refund: a 60 day cookie window, a 30 day hold period before a commission becomes payable, and a $50 minimum payout. All three can be changed per program.

Payouts we never touch

Most platforms that promise "we handle payouts" mean "we hold your money and then send it on". That requires the platform to collect funds in advance, hold a float, and often take a fee on the way through. We decided early that RelayWonder would never hold partner money. Instead, once a month the brand opens a payout batch, reviews the partners above the minimum, and exports a CSV in the exact format PayPal Payouts or Wise batch payments expect [4][5]. The brand uploads it to its own PayPal or Wise account, sends the money, and marks the batch paid.

This is less convenient than a single button, and we think it is the right trade. The brand keeps control of its cash. Partners are paid from an account they can see on the payment itself. There is no intermediary balance to reconcile, no delay while funds clear through a third account, and no reason for us to charge a percentage on the way through. If a payout is held because a fraud flag needs a human look, the money has not moved anywhere; it is still in the brand's account where it started.

The same philosophy applies to tax forms. A partner uploads a W-9 or W-8BEN once [7][8]; an operator verifies it; every brand that works with that partner sees the status. Issuing a 1099 at year end remains the paying business's responsibility, because they are the ones who paid, and we say that plainly rather than implying we take it off their hands.

Pricing that starts at zero

A program with $800 a month in partner-driven revenue should not cost $49 a month. At a 20% commission the brand is paying partners $160 and would pay the tool $49 on top, which is a 30% overhead on the program before anyone has proven it works. That is the moment most programs die, and it is the moment a free tier matters most.

Three things in that table. We are free where the competitors charge, we are roughly level in the middle, and at $50,000 we are still on a published price. We take 0% of commission on every plan, and partners never pay anything. We also do not think anyone should be impressed by this; it is simply what a SaaS partner network has to cost for a founder to try it before it has proven anything.

What we will not do

It is easier to describe a company by what it refuses. These are the lines we have drawn, and we expect to be held to them.

• We will never hold partner funds. Money moves from the brand's account to the partner's account, and we see the CSV, not the cash.
• We will never take a percentage of commission. Our price is the monthly plan, nothing else.
• We will not do Amazon. Amazon Associates has its own operating agreement and its own tracking, and nothing we build applies to it.
• We will not run agency services. We will not write your outreach for a retainer or manage your program for a cut.
• We will never show fake numbers. No invented "average partner earnings", no demo data presented as real, no survey statistics we did not run.

What comes first and what comes later

We are writing this in February 2026 and the first version is a few months away. The order matters, so here it is. First the tracking and the ledger, because a network with wrong numbers is worse than a spreadsheet. Then payouts, with the PayPal and Wise exports and the hold period logic. Then the partner side: one account, one balance per currency, the tax form upload. Then the brand side of recruiting: a program page that states the terms in numbers, invitations, and the network view of partners who have already proven themselves on similar products.

Some things are deliberately later. Searching the open web for new partners is not in the first version; the first version finds partners inside the network and from your own data. Sending outreach from inside the product is not in the first version either; drafts will be copied into your own mailbox, because we would rather you send from an address with a reputation than from ours. AI drafting and scoring of partners are ideas we like and have not built. Shopify is not supported yet; Stripe is where we start.

We will write here about what we learn, what we ship, and where we turn out to be wrong. If the recruiting gap is as real as we think, the proof will be in whether a founder with $800 in partner revenue and zero partners can get to twenty without a cold list. That is the whole bet.