Affiliate marketing

Affiliate marketing is selling someone else's product for a commission. The company that owns the product (the advertiser) pays you, the affiliate, a fixed amount or a percentage for every lead or sale you bring. You never hold stock, handle payments or answer customer calls. Your job is to find people who will buy, and prove through tracking that they came from you.

There are two very different businesses under the same name. Content affiliates run review sites, YouTube channels or newsletters, rank or build an audience, and earn commissions from links, often through retailer programmes such as Amazon's. Performance or CPA affiliates buy traffic (native, push, social, search), send it to an offer, and earn the difference between what they spent and what the offer paid. This site covers mainly the second kind, because that is where the tools, the traffic sources and the risks on this site come together.

Between the two sides usually sits a CPA network such as Dr.Cash or AdCombo. The network signs up advertisers, lists their offers, gives affiliates tracking links, counts conversions, collects money from advertisers and pays affiliates. Affiliates can work with advertisers directly, but most start with networks because networks aggregate offers, handle payments and pay smaller affiliates on a regular schedule.

As a model it is simple. As a business it is a trade: you buy clicks at one price and sell conversions at another, and the margin between them is small, volatile and entirely yours to find.

The mechanics, step by step, for a paid-traffic CPA affiliate:

1. Join a network and pick an offer. The offer page states the payout (for example "$28 per approved lead"), the allowed countries and traffic sources, the conversion event that counts, and sometimes a daily cap.
2. Get the tracking link. The network gives you a URL with your affiliate ID. You add a sub-parameter for your own click ID, so the network can tell you which click produced which conversion.
3. Buy traffic. You fund an ad network account, create a campaign, and pay per click or per thousand impressions.
4. Send the visitor through your funnel. Ad, then your prelander, then the offer page. A tracker records each click and redirects it.
5. Conversion and confirmation. The visitor converts on the offer page. The network records it and, where the offer requires it, waits for the advertiser to confirm it (in nutra COD, a call centre confirms the order).
6. Postback and payout. The network pings your tracker with the conversion and its status. At the end of the payment period the network pays you for approved conversions.

A worked example (illustrative numbers). A COD nutra offer pays $28 per approved lead. You buy 10,000 push clicks at $0.05 each: $500 spent.

• 15% of visitors click from the prelander to the offer page: 1,500 visits.
• 4% of those submit the order form: 60 leads. Cost per lead: $500 / 60 = $8.33.
• The call centre approves 35%: 21 approved leads.
• Revenue: 21 x $28 = $588. Profit: $88. ROI (profit divided by spend): 17.6%.
• EPC (earnings per click): $588 / 10,000 = $0.0588. This is your break-even bid: at a CPC above about 5.9 cents the same funnel loses money.

Now change one number you do not control. If approval drops from 35% to 25%, you get 15 approved leads, $420 revenue, and a $80 loss on the same traffic. The effective payout per lead fell from $9.80 to $7.00 while your cost per lead stayed at $8.33.

That is the whole model in miniature. You make money by improving the steps you control (click cost, prelander CTR, form conversion, choice of placements) faster than the steps you do not control (approval, payout changes, competition) move against you.

Paid-traffic affiliate marketing fits when you are comfortable with measured risk and short feedback loops: you spend today, learn within days, and keep only what the numbers support. It suits people who like analysis and testing more than building a brand, and who have a test budget they can lose without harm.

It is the wrong choice in several common situations:

• You need predictable income soon. Most first campaigns lose money while you learn. A budget that must pay rent is the wrong budget.
• You cannot tolerate the ethics of the offers available. In verticals like nutra, many profitable offers depend on claims you may not want your name next to. If that bothers you, content affiliate work with reputable programmes is a better fit.
• Your product idea is better as your own product. If you already have an audience and a product they want, the affiliate margin is the smaller share of the value you create.
• You want to build a lasting asset. A paid campaign stops earning the moment you stop paying. SEO sites, email lists and your own brands compound; bought traffic does not.
• Your traffic source forbids the offer. On Meta and on the stricter native networks, many affiliate offers simply cannot run compliantly.

A reasonable first month, assuming you already understand the terms above:

Pick one vertical and one traffic source. Nutra COD on push, or nutra on native, are common entry points because networks accept beginners and minimum deposits are low. RichAds, for example, states a $150 minimum deposit and push CPCs from $0.003 (RichAds FAQ, as of September 2026).

Join one or two networks. Expect an application form and sometimes a call with an affiliate manager who will ask about your experience and traffic. Be honest: managers can steer beginners to offers with stable approval and to geos with cheaper traffic. Ask directly what the offer's real approval is on your traffic source, and whether prelanders are provided.

Set up a tracker before your first click. A cloud tracker's free trial or a self-hosted licence on a small server both work. Test the postback with a test lead so you know conversions will arrive.

Launch one offer, one geo, several creatives. Keep everything else constant so you learn from what changes. Budget for a test of at least a few times the payout per variant.

Keep a log. Every launch, change and result, with dates. After a month, the log is worth more than the money you made or lost.

When choosing the first network and offer, favour a network whose manager answers questions and an offer whose payout, approval and call centre hours you can see before you send traffic. In the first month you will need both the help and the transparency.

The same mistakes appear again and again in beginners' first campaigns:

• Choosing the offer with the highest payout. A high payout often signals low approval, a hard geo or a saturated product. Compare effective payout per lead.
• No tracker, or a tracker without postbacks. Without conversion data per placement you cannot optimise, only switch campaigns on and off.
• Testing too many things at once. Five offers, three geos and ten creatives on a $300 budget produce noise, not answers.
• Stopping too early or too late. Killing a campaign after 20 clicks proves nothing; running it for a week without a decision rule burns money. Set the stop-loss before you launch.
• Ignoring terms. Running an offer on a traffic source it forbids, or with claims the advertiser forbids, gets conversions rejected and accounts closed.
• Not disclosing. Content affiliates who hide commissions break the FTC's endorsement guidance, which asks for a clear disclosure such as "I get commissions for purchases made through links in this post" (FTC).
• Treating a lucky week as a system. Profit that does not survive a second week of data is variance.

A handful of numbers tell you whether affiliate campaigns are working:

• ROI: (revenue minus spend) divided by spend. The final score. Positive ROI after approvals settle is the only proof that counts.
• EPC: earnings per click. Compare it with your CPC; EPC above CPC means the funnel pays for its traffic. EPC also lets you compare offers across networks.
• CR: conversion rate from click to lead, often split into prelander CTR and offer page conversion to find the weak step.
• Approval rate: approved conversions divided by all conversions. Watch it per source; sudden drops point at traffic quality.
• CPA or CPL: what one conversion costs you. Compare with effective payout, not headline payout.
• Cash cycle: days between spending on traffic and receiving the payout. Low ROI with a fast cycle can beat high ROI with a slow one.

Read all of these per placement and per creative, not only per campaign, or averages will hide both your winners and your losers.

The risks of the model are mostly structural, not technical.

You do not own the customer. The advertiser keeps the buyer, the data and the repeat sales. If the offer is paused, your campaign stops with it.

Payout and approval can change overnight. Networks cut payouts, advertisers tighten confirmation, caps fill. Diversify across offers and networks.

Counterparty risk. Networks go under or delay payments. A pattern of "technical delays" is a warning sign; so is a network that cannot explain rejected conversions.

Shaving. Some networks or advertisers under-report conversions. It is hard to prove without your own tracker comparing leads sent with leads counted.

Legal and reputational exposure. Regulators treat affiliates who spread deceptive claims as participants, not bystanders. The ad account ban is the fast consequence; the legal one is slower and larger.

Platform dependency. A single traffic source's policy change can wipe out a working campaign in a day.