Back to overview
Lesson 4 of 6

Tracking and Attribution for Ecommerce

8 min read

The Attribution Challenge in Ecommerce

Ecommerce attribution is more complex than in other verticals because customer journeys span multiple devices, sessions, and touchpoints over days or weeks. A customer might discover a product through an affiliate's Instagram post on their phone, research it on a laptop via the affiliate's blog review, and complete the purchase three days later through a direct visit on their tablet. Correctly attributing this sale to the affiliate who initiated the journey -- and not to the last coupon site the customer visited before checkout -- is the central technical challenge.

The stakes are real. Incorrect attribution either overpays affiliates for sales they did not influence (eroding margins) or underpays affiliates for sales they did drive (causing top partners to leave for competitors with more accurate tracking). Both outcomes damage program economics.

Tracking Methods Comparison

MethodHow It WorksAccuracyLimitations
First-Party CookiesAffiliate click sets a cookie on the merchant's domain; cookie is read at checkoutMedium-HighBlocked by Safari ITP (7-day cap), Firefox ETP; does not work cross-device
S2S (Server-to-Server) PostbacksClick ID stored server-side; conversion posted back to tracking platform via APIHighRequires deeper integration; does not help with impression-based attribution
Coupon Code AttributionUnique codes assigned to each affiliate; code entered at checkout triggers attributionHigh (for code users)Not all customers use codes; can be leaked to unauthorized sites
Sub-ID / UTM TrackingCustom parameters appended to affiliate links; captured at conversionMediumDepends on URL parameter persistence through the shopping flow
Fingerprint / ProbabilisticDevice characteristics used to match clicks to conversions when cookies failLow-MediumPrivacy regulations (GDPR, CCPA) restrict usage; accuracy degrades with scale
Cross-Device GraphsDeterministic (logged-in) or probabilistic matching across devicesMedium-HighRequires user login or third-party identity graph; adds cost and complexity

Cookie Windows and Their Impact

The cookie window (or attribution window) defines how long after an affiliate click a conversion can be credited to that affiliate. In ecommerce, this window must match the product's typical consideration period. A $30 t-shirt might have a 2-day consideration cycle, while a $2,000 mattress could take 30-45 days from first research to purchase.

  • Impulse/low-AOV products ($10-50): 7-14 day cookie window is typically sufficient
  • Considered purchases ($50-200): 30-day window captures the research-to-purchase cycle
  • High-AOV/complex purchases ($200+): 45-90 day windows may be needed for products with long sales cycles
  • Subscription products: 30-day click window plus first-renewal attribution for recurring commission models
  • Safari ITP caps first-party cookies at 7 days for script-set cookies -- S2S tracking bypasses this limitation

Browser privacy changes are steadily reducing cookie reliability. Safari ITP, Firefox ETP, and Chrome's evolving privacy sandbox all affect cookie-based tracking. Any ecommerce affiliate program that relies solely on client-side cookies will see increasing attribution gaps. S2S postback integration should be a priority for programs running at scale.

Coupon Code Attribution Strategy

Coupon codes are a powerful attribution method in ecommerce because they work regardless of device, browser, or cookie status. When a customer enters an affiliate's unique code at checkout, the attribution is deterministic -- no probabilistic matching or cross-device reconciliation needed. This makes coupon codes particularly effective for influencer and social media affiliates whose audiences discover products through content that does not always include trackable links.

The challenge is code leakage. Coupon aggregator sites scrape and redistribute affiliate codes, claiming attribution for organic customers who simply searched for a discount before checking out. To mitigate this, operators can use single-use codes, time-limited codes, or codes that are only valid when paired with the affiliate's tracking link.

Multi-Touch Attribution Considerations

Most ecommerce affiliate programs use last-click attribution: the affiliate whose link was clicked most recently before purchase gets full credit. This model is simple but creates perverse incentives -- coupon and deal sites that intercept customers at checkout claim credit for sales initiated by content creators higher in the funnel. More sophisticated programs are moving toward first-click priority (crediting the affiliate who introduced the customer) or split attribution (dividing commission between the first and last touchpoint).

If your program uses last-click attribution, consider implementing a "content creator override" rule: when a content affiliate's cookie exists and a coupon affiliate's click occurs within 60 minutes of checkout, the content affiliate retains attribution. This protects the partners who actually drive discovery and purchase intent.

Key Takeaways

  • S2S postback tracking is more reliable than cookie-based methods as browser privacy restrictions increase
  • Cookie windows should match the product's actual consideration period -- not an arbitrary industry default
  • Coupon code attribution works cross-device and cross-browser but requires leakage prevention measures
  • Last-click attribution incentivizes coupon poaching -- consider first-click priority or content creator override rules
  • Cross-device tracking requires either logged-in user matching or probabilistic identity graphs, each with trade-offs