Keep a creator eligible for credit long enough to cover the purchase delay you intend to reward, within the limits of your tracking system. Compare candidate windows against observed click-to-order times before choosing one. An affiliate attribution window sets payment eligibility; it does not establish how much the creator caused the purchase.
A seven-day window may fit a program built around purchases soon after a recommendation. A considered purchase may justify testing thirty days or longer. Those are candidate policies, not industry rules. The decision needs a defined starting event, a credit-allocation rule, and evidence that your system can apply both.
Define which clock you mean
The click window answers one question: how long can a recorded referral remain eligible for attribution?
Rewardful's documentation defines its cookie window in those terms. It documents a sixty-day default that merchants can configure by campaign. That is a Rewardful setting, not a recommended duration for every business.
Keep these other clocks separate:
| Clock | Question it answers |
|---|---|
| Click eligibility | Can this referral still receive credit? |
| Reporting range | Which dates appear in this report? |
| Commission review | When do returns and cancellations stop delaying approval? |
| Payout schedule | When does approved money reach the creator? |
For example, Shopify Collabs performance reporting displays the last thirty days by default. Changing that report range does not establish a thirty-day referral window.
Also separate eligibility from the winner. Two creators might have eligible clicks for one order. Your first-click, last-click, or other allocation rule decides who receives credit. Settle how links and discount codes share credit before comparing window lengths, so the analysis does not mix two policy changes.
Measure the delay you can observe
Start with orders that have a recorded click and purchase timestamp. Record the order ID, creator, click time, purchase time, eligible order value, attribution method, and refund status. Use one timezone and retain timestamps rather than rounded dates.
For each order:
Delay in days = (purchase timestamp - qualifying click timestamp) / 86,400 seconds
The qualifying click must follow your chosen allocation rule. If you use last eligible affiliate click, measuring from the first visit gives a different answer.
Group results by product or purchase process where it matters. A replenishment purchase and an item requiring a size check may have different delays. Keep new and returning buyers distinguishable if your program treats them differently.
Watch for a circular result. If the existing system stops recording eligible referrals after seven days, its attributed-sales report cannot reveal all purchases that happened later. A report containing only seven-day conversions will make a seven-day window look complete. Mark this measurement limit before drawing a conclusion.
Likewise, leave recent clicks time to convert. A thirty-day comparison needs a cohort whose clicks have had thirty days of observation. Keep unresolved returns separate when estimating final commission cost.
Work through short and long purchase paths
The following order-level data is hypothetical. It illustrates policy effects, not a benchmark or customer result.
Assumptions: each buyer has one recorded affiliate click, no competing code or later affiliate click, and an approved purchase. Eligible value excludes tax and shipping. No order has a refund. All delays come from exact timestamps, and the comparison counts a purchase at or before the window boundary as eligible. Your platform may implement boundaries differently.
| Order | Hypothetical path | Delay | Eligible value | 7 days | 30 days |
|---|---|---|---|---|---|
| A | Click, then buy that afternoon | 0.25 days | $40 | Yes | Yes |
| B | Click, check size, buy next day | 1 day | $60 | Yes | Yes |
| C | Click, compare options, buy later | 5 days | $100 | Yes | Yes |
| D | Click, wait for payday, buy | 12 days | $200 | No | Yes |
| E | Click, discuss purchase, buy | 24 days | $300 | No | Yes |
| F | Click, postpone purchase, return | 40 days | $400 | No | No |
Calculate order coverage as eligible orders divided by all six observed orders. Calculate value coverage as eligible value divided by the $1,100 total observed value. Neither figure is a conversion rate because the denominator excludes visitors who never bought.
Assume a hypothetical 10% commission on eligible value:
| Candidate | Eligible orders | Order coverage | Eligible value | Value coverage | Commission |
|---|---|---|---|---|---|
| 7 days | 3 | 50.0% | $200 | 18.2% | $20 |
| 30 days | 5 | 83.3% | $700 | 63.6% | $70 |
Percentages are rounded to one decimal place. Dollar calculations are exact in this example. Extending the window adds two eligible orders, $500 of attributed value, and $50 of commission. It creates no additional purchases in this fixed dataset.
If your program intends to reward the longer consideration paths represented by D and E, thirty days covers them under these assumptions. If it intends to reward only near-purchase referrals, seven days implements that narrower policy. Explain the choice to creators; do not present the added attributed value as proof of additional sales.
Complete the policy worksheet
Use this completed hypothetical worksheet as a model. Every entry is a proposed business rule, subject to platform verification.
| Decision | Example policy |
|---|---|
| Business reason | Cover considered purchases like D and E |
| Starting event | Recorded affiliate-link click |
| Window | Thirty elapsed days from the qualifying click |
| End boundary | Purchase timestamp at or before expiry |
| Competing creators | Last eligible affiliate click receives credit |
| Repeat clicks | A new valid click starts its own eligibility period |
| Code priority | No code override in this link-only example |
| Repeat orders | No recurring entitlement from the first purchase |
| Missing tracking | Review documented disputes; never promise automatic recovery |
| Changes | Apply revised rules to new clicks from the announced effective date |
Window selection cannot settle refunds or payout timing. Keep those in a separate policy for reconciling commissions after returns.
Check whether your system can honor the policy
Ask the vendor which event becomes attributed: a purchase, signup, or another event. Confirm whether later subscription payments follow a separate commission rule. Do not assume that a cookie's lifetime also defines the duration of recurring commission.
Then verify the proposed window, exact expiry boundary, repeat-click behavior, and competing-code priority. Check what happens when the purchase occurs on another device or the referral identifier is unavailable. A longer configured window cannot, by itself, supply a missing click-to-order connection.
Public setup pages may leave these details unresolved. Shopify's Collabs program instructions describe link destinations, URL parameters, discounts, and commissions. That page alone does not establish a configurable click window. Obtain confirmation for the system you use rather than filling gaps with assumptions.
Modash's affiliate-software comparison raises useful evaluation questions about link and code tracking, order-level reporting, and returns. Treat its vendor comparisons as research leads. Verify the mechanics in your own integration.
Before launch, test the affiliate links with the paths your policy promises to recognize. Include an order inside the window, one outside it, and competing referrals. Save the expected and actual recipient of credit.
Keep causal claims and disclosures separate
A timestamp shows that a recorded click preceded a purchase. It cannot show what the buyer would have done without that recommendation. Google's Conversion Lift documentation explains causal measurement through controlled comparisons between treatment and control groups. That principle explains why window-based attribution alone cannot answer the causal question. It does not mean an affiliate program has access to Google's experiment tools.
For endorsements affecting US consumers, commission eligibility also creates a disclosure issue. The FTC's affiliate-marketing guidance calls for clear, conspicuous disclosure of the relationship near the recommendation. The words "affiliate link" alone may not tell readers that purchases generate payment. Explain the earning relationship in plain language; choosing a short window does not replace that disclosure.
Choose two candidate windows, complete the worksheet, and ask your tracking owner to verify each rule before you announce the terms to creators.
