How Affiliate Cookies Actually Work
— 30 vs 90 Day

If you’ve been around affiliate marketing for any time, you’ve heard the advice to “look at cookie duration.” Most articles treat that as a single number to compare. It isn’t. The cookie is the program telling you how long your work persists — and whether 30 days is enough depends entirely on what you’re recommending.

This article pulls cookie mechanics apart using three programs we’re actively running: Amazon (24-hour), Shopify (30-day), and WunderTools (90-day). Same vertical traffic profiles. Different cookie math. The difference in captured commissions is large — and predictable.

What an Affiliate Cookie Actually Is

An affiliate cookie is a tiny piece of data the merchant’s system drops into the visitor’s browser when they click your tracking link. It contains three things: your publisher ID, the timestamp of the click, and the program’s attribution window. Nothing more.

When the visitor converts — signing up, buying, completing a trial — the merchant’s system checks the browser for that cookie. If a valid cookie exists and the conversion is within the window, your publisher ID is attached to the sale and you earn the commission.

That’s the entire mechanism. A cookie isn’t a tracking pixel, isn’t a fingerprint, isn’t a permanent identifier. It’s a single browser file with an expiration timer attached to one publisher ID.

How Attribution Flows From Click to Commission

The actual flow is more involved than “click → cookie → payout.” Five players are involved, and each step is a place where attribution can break.

1. Click. A visitor lands on your article and clicks your affiliate link. The link is something like https://merchant.com/?aff=YOUR_ID — your publisher ID is embedded in the URL.
2. Redirect through the tracker. The merchant (or its network) receives your request, records the click with your publisher ID against a timestamp, and serves a redirect to the merchant’s actual page. The redirect is where the cookie gets dropped.
3. Browse phase. The visitor browses the merchant’s site. They may or may not buy today. They may bounce and come back tomorrow via Google. They may bookmark and return Saturday.
4. Conversion. At some point the visitor converts. The merchant’s system re-reads the cookie file in their browser. If it exists and the timestamp + cookie window is still valid, your publisher ID is attached to the conversion.
5. Reporting + payout. The merchant (or network) credits the commission to your account. The commission enters your dashboard but isn’t payable yet — it has to clear the program’s locking window (typically 30–60 days) to wait for refunds and cancellations.

Every step is a potential failure point. Redirect hijacking drops your publisher ID in step 2. Browser settings clear cookies between steps 3 and 4. Post-cookie attribution by the merchant’s own marketing email can override your cookie in step 4. These are real failure modes we’ve seen — covered in more detail below.

But the variable you control — the one you pick at application time — is the window in step 4. That’s the only piece the program commits to up front.

The 30 vs 90 Day Math, With Real Numbers

Consideration window is the amount of time your reader takes between reading your article and buying. It varies by product category, audience intent, and price point. A 24-hour cookie covers impulse purchases. A 30-day cookie covers most SaaS research phases. A 90-day cookie covers long research and comparison phases that involve more than one reading session.

The math problem is straightforward. If your reader’s decision window is longer than the cookie, you lose commissions you actually drove.

This is why “30 vs 90 day” is not a marketing detail. The difference between a 30-day and 90-day cookie on the same article, with the same traffic, can easily mean 30–50% more captured commissions — purely because of the attribution math, with no additional work on your end.

Three Live Programs We Run, Side by Side

Here are three programs running in our stack right now, with the cookie windows we’re dealing with. The point isn’t to crown one program better than another — it’s to show how a cookie window interacts with realistic traffic patterns.

Honest take: Amazon’s 24-hour cookie is the industry’s most-discussed short window. It’s brutal for any product where readers deliberate: monitors, chairs, premium electronics — anything where people bookmark and come back days later. The 24-hour window was built for the Amazon shopping flow where the user adds to cart and checks out in one session. For affiliate review content with multi-day decision windows, it’s a known conversion loss. Our Amazon earnings are real but always smaller than the click volume suggests — because a meaningful slice of those clicks come back on day 3, day 5, day 8.

Honest take: 30 days is the SaaS industry standard for good reason. Shopify buyers typically research for 1–2 weeks, compare against WooCommerce and BigCommerce, talk to their co-founder, and convert. A 30-day window covers that flow cleanly. The trade-off: if your reader’s research phase stretches past month-end — because of a Q4 freeze, a vacation, a launch delay — you lose attribution. We see it: a Shopify click on day 1 that converts on day 38 is credit to direct traffic, not us.

Honest take: 90 days is generous and unusual. For AI productivity tools specifically, it’s the right window — buyers evaluate against ChatGPT Team, Notion AI, and a handful of alternatives over weeks, install trial versions, bring in teammates. A 30-day cookie would lose every multi-stakeholder conversion. The 90-day window means our WunderTools article captures conversions that other programs’ articles on the same product would lose. The cost: the program has to absorb a longer “in escrow” window on its books. Not every merchant will.

Three programs. Three cookie windows: 1 day, 30 days, 90 days. The 30-day and 90-day gaps look small numerically, but they’re capturing fundamentally different research behaviors across the readers we send them.

Cookie Traps That Lose Commissions You Drove

Cookie duration is the headline statistic. What hides inside the cookie policy is at least as important. Four traps we’ve personally lost commissions to:

The honest summary: the cookie headline is necessary but not sufficient. Cookie policy, cross-device tracking, and post-cookie override rules decide what fraction of your real conversions actually show up in your dashboard.

Pick Programs by Your Reader’s Consideration Window

Tie the cookie window to your reader’s decision pattern, not to the merchant’s headline. Quick reference by category:

If the cookie at the program you’re evaluating falls short of this range for the reader you’re sending, you’ll do the work and lose the commission. That’s the silent killer of niche affiliate sites — not bad content, but the cookie expiring before the reader decides.

What to Do This Week

Three concrete moves, ranked by leverage:

1. Audit your top 5 articles’ cookie windows. Open every affiliate program you promote. Find the cookie window in your partner dashboard or the program’s TOS. If any article’s window falls below the category’s consideration range above, find a peer program with a longer window — even if the headline commission rate is lower.
2. Document the cookie policy in your spreadsheet. Beyond “30 day”, write down: does this program override cookies with merchant email? Does it last-click to paid search? Does it track across devices? The headline is a starting point; the policy is the actual ceiling.
3. Add cookie language to your CTAs. A line like “Link sets a [X]-day cookie so you can compare before buying” helps readers understand they don’t have to rush — and improves conversion on the readers who would otherwise have abandoned the cart. The cookie helps you, but readers don’t know that unless you tell them.

Cookie duration is one of those mechanics that sounds trivial until you trace the lost revenue. Then it’s the single most leverage-rich variable in the program — because changing it doesn’t require better content, more traffic, or new links. It just requires picking programs whose window fits your reader.