Affiliate Marketing: A Practical Playbook
Affiliate marketing is simple to explain and slow to get right: you help someone make a buying decision, link to the product, and earn a commission if they buy. This playbook covers the parts that actually move the needle.
Numbers in this post are labelled as typical ranges or as program terms you should re-check. Programs change their terms often; always read the current agreement before you build on it.
1. Choose a niche with buying intent
The best niches are ones where people actively research before spending money. Passion helps, but intent pays.
Good signals
- Considered purchases: products people compare before buying (software, outdoor gear, coffee equipment, home office, cameras).
- Comparison searches exist: Google autocomplete shows “best …”, “… vs …”, “… review” for your topic.
- Multiple monetizable products: at least 10–20 products or tools you could genuinely recommend.
- Recurring or high-ticket options: SaaS subscriptions or pricier hardware make each conversion worth more.
- You can add real experience: hands-on testing, photos, data. This is your moat against generic content.
Red flags
- Results pages dominated by huge publishers and the brands themselves for every query you check.
- Only low-priced items with low commission rates.
- Topics where advice can cause harm (health, finance, legal) unless you have real credentials. These face stricter quality expectations.
Quick test: write down 30 keyword ideas in the niche. If you can’t find 30 with clear commercial intent, the niche is probably too thin.
2. Pick your programs
You’ll usually mix a broad retailer program, one or two networks, and direct programs from the best brands in your niche.
| Program | What it is | Commission & cookie (check current terms) |
|---|---|---|
| Amazon Associates | Retail program covering almost any physical product. | Category-based rates published in Amazon’s fee schedule, typically low single digits up to about 10% for most categories. Standard cookie is 24 hours (longer, up to 90 days, if the item is added to cart). Easy to start, low per-sale payout. |
| Impact (impact.com) | Network/platform used by many large brands and SaaS companies. | Set by each advertiser. Varies widely: percentage of sale, flat fee per sale or lead. Cookie windows are commonly 7–90 days, depending on the brand. |
| Awin | Large network, strong in Europe (incl. DACH) and the UK. | Set by each advertiser. Retail commonly low-to-mid single-digit to double-digit percentages; cookie lengths vary by program (often 30 days). |
| CJ (Commission Junction) | One of the oldest networks; many big retail, travel and finance brands. | Set by each advertiser; percentage or fixed fee. Cookie windows vary per program. |
| PartnerStack | Network focused on B2B SaaS. | Set per SaaS company; recurring commissions (e.g. a share of subscription revenue for 12 months or for the customer’s lifetime) are common. Cookie windows often 30–90 days. |
| Direct SaaS programs | Run in-house by the company (often via Rewardful, FirstPromoter, Tolt etc.). | Recurring percentages and long cookies are common in SaaS, but terms differ widely. Read the fine print on attribution and payout thresholds. |
How to compare programs
- Earnings per click (EPC) matters more than headline commission. A 4% program that converts well can beat a 30% program nobody buys from. Many networks show EPC per advertiser.
- Cookie length and attribution: last click? first click? Does a coupon site override you?
- Recurring vs one-time: recurring SaaS commissions compound over time.
- Payout threshold, currency and timing: relevant if you’re in Switzerland and paid in USD or EUR.
- Restrictions: brand-keyword bidding bans, coupon rules, email restrictions, country limits.
3. Keyword research for “best X”, “X vs Y” and “X review”
These three patterns capture people near a decision. Start with them before you write any informational content.
| Pattern | Searcher’s state | Example |
|---|---|---|
| best X (for Y) | Knows the category, needs a shortlist | best standing desk for small spaces |
| X vs Y | Down to two options | notion vs obsidian |
| X review | Validating one choice | framework laptop 13 review |
| X alternatives | Unhappy with or priced out of X | mailchimp alternatives |
Process
- Seed list: list products, brands and categories in your niche.
- Expand: run each through Google autocomplete, “People also ask”, and a keyword tool (Ahrefs, Semrush, or free options like Google Keyword Planner and Search Console once you have traffic).
- Modifiers: add “for beginners”, “under $500”, “for small apartments”, “2026”. Long-tail, specific queries are easier to rank for.
- Check the SERP manually: if forums, small blogs or thin pages rank in the top 10, you have a chance. If it’s all major publishers, move on for now.
- Prioritize: intent first, then difficulty, then volume. A 200-searches-a-month “X vs Y” page can out-earn a 10,000-searches informational page.
4. Page types that convert
- Roundups (“best X”): a quick-picks summary at the top, then detailed sections, a comparison table, and how you tested.
- Comparisons (“X vs Y”): verdict up front (“Pick X if…, pick Y if…”), side-by-side table, then deep dive.
- Single reviews: who it’s for, who should skip it, real pros/cons, photos or screenshots, pricing, alternatives.
- Alternatives pages: convert well for SaaS because the reader is already looking to switch.
- Tools and calculators: e.g. a cost calculator that recommends a product. Harder to build, strong to rank and link.
Elements that consistently help
- Clear verdict above the fold and a visible call to action.
- Evidence of first-hand use: original photos, test results, screenshots.
- Honest downsides. They build trust and therefore clicks.
- Updated date and a short “how we evaluate” section.
- Fast, readable pages on mobile: most affiliate traffic is mobile.
5. Setting up the site (Next.js on Vercel)
You don’t need a heavy stack. A static or mostly static site is fast, cheap and SEO-friendly.
- Framework: Next.js with static export or static generation; write posts in MDX or pull them from a headless CMS.
- Hosting: Vercel with a Git-connected repo: every push deploys and every pull request gets a preview URL.
- Performance: optimized images (sized, lazy-loaded, modern formats), minimal client-side JavaScript, system or self-hosted fonts. Aim to pass Core Web Vitals.
- SEO basics: unique titles and meta descriptions, one H1, a clean URL structure (
/best/standing-desks), canonical tags, an XML sitemap,robots.txt, internal links between related pages. - Structured data:
Articleand, where appropriate,Product/Reviewschema. Only mark up what is actually on the page. - Analytics: privacy-friendly analytics (e.g. Vercel Web Analytics, Plausible) plus Google Search Console.
- Launch hygiene: keep
noindexon while you build (like this site), then remove it deliberately when the first pages are ready.
6. Disclosure rules
Not legal advice. This is a general overview. Rules depend on your country, your audience and the specific situation; consult a qualified lawyer for your setup.
United States: FTC
- The FTC’s Endorsement Guides (16 CFR Part 255, updated 2023) require disclosing a “material connection”, which includes earning a commission.
- Disclosures should be clear and conspicuous: close to the recommendation or link, before the reader clicks, in plain language (“I earn a commission if you buy through these links”). A footer-only note or a separate disclosure page alone is generally not enough.
- Amazon Associates additionally requires a specific statement, e.g. “As an Amazon Associate I earn from qualifying purchases.”
Switzerland and the EU: general principles
- Switzerland: the Federal Act against Unfair Competition (UWG/LCD) prohibits misleading business practices; commercial communication should be recognizable as such. The Swiss Fairness Commission’s principles also expect advertising to be identifiable.
- EU: the Unfair Commercial Practices Directive (2005/29/EC, as amended) treats disguised advertising as misleading; e.g. paid content presented as editorial without making that clear. Member states implement this in national law, and some (e.g. Germany) have strict case law on labelling advertising.
- Reviews: under EU rules, if you claim reviews come from real users, you should be able to say how you ensure that. Never fake reviews or ratings.
- Practical takeaway: label affiliate links and sponsored content clearly and near the link, don’t overstate claims, keep prices and availability accurate or dated, and comply with privacy/cookie rules (GDPR, Swiss revFADP) for tracking and email.
7. Link tracking and UTMs
- Use sub-IDs: most networks let you append a sub-ID (e.g. Impact
subId1, Awinclickref, Amazon tracking IDs per page/section). Use them to see which page and which button earned the sale. - Clean redirect links: route affiliate links through your own path (e.g.
/go/product-name) so you can update destinations in one place. Mark themrel="sponsored nofollow". - UTMs are for your traffic sources, not affiliate links: tag newsletter, social and partner links with
utm_source,utm_medium,utm_campaignto see which channels bring readers who click. - Track outbound clicks as events in analytics; compare click-through rate per page against network-reported conversions.
/go/acme-desk → https://partner.example.com/acme?subId1=best-standing-desks-table
newsletter link → https://mutz.mutz.dev/best/standing-desks?utm_source=newsletter&utm_medium=email&utm_campaign=oct-20268. Email capture
Search traffic is rented from Google. An email list is yours.
- Offer something specific: a buying checklist, a comparison spreadsheet, a deals digest. “Subscribe to our newsletter” rarely converts.
- Place forms where intent is high: end of reviews, inside comparison pages, as an inline box, not just the footer.
- Automate a welcome sequence: 3–5 emails delivering the lead magnet, your best recommendations and how you test.
- Stay compliant: clear consent, double opt-in is wise in the EU/CH, easy unsubscribe, privacy policy. Check whether each affiliate program allows links in email. Some don’t (Amazon restricts this).
9. Realistic timelines and numbers
These are typical ranges commonly reported by practitioners, not guarantees or measured statistics. Your results depend on niche, competition, content quality and luck.
| Phase | Typical range | What’s happening |
|---|---|---|
| Months 0–3 | Little to no search traffic | Publishing, getting indexed, building first links. Many new sites see few clicks in this phase. |
| Months 3–6 | First rankings on long-tail queries | A few pages start to pull traffic; first commissions are possible but small. |
| Months 6–12 | Steady growth if content and quality hold | Double down on what ranks; update and expand winning pages. |
| 12+ months | Meaningful income possible, not assured | Compounding content, links and email. |
Do the math with your own inputs
Estimate monthly earnings as visitors × click-through rate to merchant × merchant conversion rate × average commission. Plug in conservative guesses, then replace them with your real data as soon as you have it. For example, if you assume 5,000 visitors, a 10% click-through, a 3% conversion and a 5 CHF commission, the result is 75 CHF/month. The point is to see which lever matters most, not to predict income.
10. Your 30-day action plan
Week 1: Foundation
- Pick a niche and validate it with 30+ commercial-intent keywords
- Shortlist 10–20 products you can test or have used
- Apply to 3–5 relevant programs (one network, one retailer, direct brand programs)
- Register a domain and set up Next.js + Vercel with noindex on
Week 2: Structure
- Map 15–20 target keywords to page types (roundup, vs, review, alternatives)
- Build page templates: comparison table, verdict box, pros/cons, disclosure line
- Write About, How we test, Affiliate disclosure and Privacy pages
- Set up analytics, Search Console, sitemap and outbound click tracking
Week 3: Content
- Publish 3–5 bottom-of-funnel pages with original photos or screenshots
- Set up
/go/redirect links with sub-IDs per page - Create a lead magnet and an email signup form with a welcome sequence
Week 4: Launch and iterate
- Remove noindex, submit the sitemap, request indexing for key pages
- Publish 2–3 more pages and interlink everything
- Share genuinely useful pages in 2–3 relevant communities (follow their rules)
- Set a monthly review: rankings, clicks, conversions, pages to update
11. Common mistakes
- Choosing programs by headline commission instead of EPC and fit.
- Writing informational content first and never getting to the pages that earn.
- Generic, rewritten content with no first-hand experience. It’s hard to rank and harder to trust.
- Recommending everything: if every product is “great”, nothing is.
- Hidden or vague disclosures.
- No tracking: not knowing which page or button earns.
- Relying on one program or one traffic source: commission cuts and algorithm updates happen.
- Never updating: stale prices, discontinued products and old dates erode rankings and trust.
- Quitting at month three, right before results usually begin.
Disclosure: this post contains no affiliate links. If future posts on mutz lab do, they’ll be labelled clearly next to the link.