Class 74 taught you to test paid creative you control and pay for directly. Affiliate distribution works differently: someone else recommends your product — or you recommend someone else's — and a tracked trail connects the recommendation to a completed action and a commission. When the fit is real, it is efficient. When the fit is fake, no tracking code saves it.

The flow, end to end

Memorize the chain before touching any platform:

useful content → relevant audience → clear recommendation
→ tracked destination → completed customer action → attributed commission

Each arrow does distinct work. Useful content earns attention around a real question. A relevant audience means the question belongs to them. A clear recommendation names the product, the job it does, and its boundary. A tracked destination carries the visitor to the right page without leaks. A completed action — purchase, signup, booked trial — is the merchant's promise fulfilled. Attribution connects that action back to the recommender so the commission is fair.

Learn the roles plainly:

• Merchant (advertiser): owns the product, fulfills the order, pays commission, handles support and refunds.
• Affiliate partner (publisher, creator): owns the audience relationship and the recommendation. Earns commission only when the tracked action completes.
• Offer: the product plus terms — price, trial, refund, cookie or attribution window, commission rate, and what counts as a conversion.
• Link or code: the tracked identifier that ties a visit or purchase to a partner.
• Attribution: the rulebook for assigning credit when several touches exist.
• Conversion and payout: the verified action and the schedule on which the merchant pays.

A code does not transform an irrelevant offer into value. If a finance tutorial audience suddenly hears a pitch for an unrelated gaming headset because the commission is high, the chain breaks at the second arrow. The audience notices, trust dips, and the few clicks that survive rarely convert because the destination never matched the need.

Why trust comes before tracking

Beginners shop for affiliate programs by commission rate. Experienced builders shop by recommendation logic: could I explain this product to my specific audience, from firsthand or well-sourced experience, with a straight face and a clear boundary?

Run the three running examples through that filter. Research Desk fits a creator who teaches source-linked research workflows — the audience already asks "how do I verify claims from filings?" and the product answers exactly that. The tailor fits almost no affiliate logic at all; a local service earns through referrals and map discovery, not coupon links. ByeBuy Classroom could recommend a research or hosting tool it genuinely uses in lessons, with proof on screen — but only the tool it would name even without a commission.

That last test is the whole ethic: recommend what you would recommend for free, then disclose and track the paid version cleanly. Part XIII's creator-integrity rule carries directly into this class. The audience remembers a bad recommendation far longer than you remember the payout.

Platform mechanics are secondary and change often — networks, dashboards, payout thresholds, approval flows. Treat them as a refreshable tool layer. What endures is the logic: real need, honest explanation, working destination, verified action, fair credit.

Exercise: describe the fit before the forecast

Create AFFILIATE-FIT.md for one product and one audience. No revenue projections — this doc proves the recommendation deserves to exist. Include:

• Audience problem: the recurring job in their words. Example: "turn a 100-page filing into a citable brief without inventing quotes."
• Genuine reason to recommend: your firsthand or sourced basis. Used it, tested it, verified its docs, interviewed users — state which.
• Proof: the artifact you can show — screen capture, sample output, support transcript, refund policy screenshot.
• Recommendation boundary: who should not buy it and what it does not do. Honest limits increase conversion among the right people.
• Link destination: exact page, not the homepage — the sample brief, pricing, or setup guide that continues the promise.
• Action and commission: what counts (trial start, paid plan, booking) and the rate and window, quoted from the program terms.
• Fit sentence: "We recommend [product] to [audience] for [job] because [evidence], except [boundary]."

Example — creator-led research publication recommending Research Desk:

Finish line: an AFFILIATE-FIT.md with a recommendation logic a skeptical reader would accept — problem, evidence, boundary, destination, and action.

Verify quickly: cover the commission line and read the rest. Would the recommendation still make sense if the commission were zero? If not, change the product, not the paragraph.

Common failure mode: a fit doc that is really a revenue fantasy — "10,000 followers times 2% times 50 EUR" — with no evidence anyone needs the product. Math never fixes mismatch.

Check your understanding

1. Recite the six-step affiliate flow and name what breaks if the audience is irrelevant. 2. What is the difference between a merchant and an affiliate partner, and what does each owe the customer? 3. Why write the recommendation boundary before any forecast?

Next

You understand the path. Lesson 75.2 teaches selection discipline — scoring candidate offers by need, trust, proof, quality, and destination so you choose one and reject two with reasons.