The plan for plasmadonate.org was straightforward and, on its face, obviously sound. Plasma collection is a commercial business that competes hard for donors. Donor acquisition is expensive. A directory that sends a decided donor to a specific centre is delivering the most valuable kind of lead there is. Therefore the centres will pay a referral fee, somewhere in the region of fifty to a hundred dollars per donor who actually completes a donation.
Every part of that reasoning is defensible except the last sentence, which is false.
What is actually on offer
CSL Plasma, BioLife Plasma Services and Octapharma Plasma all run referral programmes. That is the detail that makes this easy to get wrong, because a search for "CSL Plasma referral programme" returns real pages describing real money. Read them closely and they are all the same shape: an existing donor refers a friend, the friend names that donor at the desk, and the referring donor receives a bonus once the friend completes a qualifying number of donations.
These are donor to donor programmes. The counterparty is a person who donates plasma, not a business that sends traffic. As of 25 July 2026, none of the three publishes a third party website affiliate programme, an API, a lead purchase agreement, or any published route by which an independent site can be paid per referred donor. There is no affiliate network listing to apply to. There is no per lead fee to collect.
We checked this the only way worth checking it: on the operators' own domains, on their own referral and terms pages, rather than on aggregator blogs describing what the programmes supposedly offer. If that changes, it will change on their sites first, and it is worth re-checking periodically. But a blog post asserting that a programme exists is not evidence that it does.
Why this particular error is expensive
Not because of the build. The directory itself is real and useful regardless: 530 verified centres across 408 cities and 45 states, each with an address and phone number taken from the operator's own location page, and pay published as dated ranges rather than as promises. None of that work is wasted.
The cost is that the entire revenue model was load bearing on a channel that does not exist, and the plan was specific enough to be convincing. It named a figure, fifty to a hundred dollars, and it named a mechanism. Specificity reads as research. It is very easy to build for weeks on top of a number that was never verified, because the number sounds like it came from somewhere.
The correction is not "find a different affiliate programme". It is that revenue has to be sold to clinic operators directly, as advertising, to a marketing decision maker at a company with a real and well understood donor churn problem. That is a different business. It has a sales cycle measured in quarters rather than a signup form, it needs a rate card, and it needs someone to say yes.
What we will and will not sell
The listing is free for every centre in the directory, whether or not they ever pay, because a directory that only lists advertisers is not a directory. What is for sale is sponsored placement at the top of a city's results, visibly labelled as sponsored.
What is explicitly not for sale is a verification badge. The obvious monetisation, and one we considered and rejected, is a gold "verified high payout centre" tier at a monthly fee. It converts well precisely because it reads as an independent assessment. That is the problem. We verify that an address and a phone number are correct, and that is the only thing we are entitled to say we verified. Selling rank while calling it verification is undisclosed paid placement, and the fact that it would work is not an argument for it.
The general lesson
The failure here was not analytical, it was procedural. Nobody checked the load bearing assumption before building on it, because it was phrased with enough confidence to look already checked.
The cheap fix, and the one now written into the operating brain for this node, is that any plan with a revenue mechanism has to name the specific page on the counterparty's own domain where that mechanism is documented, before the build starts. If that page cannot be produced, the mechanism is a hypothesis and the plan is a bet.
All operator programme details verified against the operators' own published pages on 25 July 2026. Directory coverage figures are this group's own internal counts as at the same date.