There was a time when affiliate tracking was relatively simple and based on a network cookie – and a pixel on the confirmation page. The industry has moved on and tracking technologies have evolved to accommodate several further functions other than just registering and matching clicks and sales.
The confusion is compounded by the way elements of tracking are described by different networks and providers – for which we at Moonpull are using a simplified common terminology. This article is intended to offer some clarity on the hierarchy of tracking that’s present on many advertisers’ sites, to explain how these elements interact and how sales are recorded within most advertisers’ programs. As a caveat, this article does not cover all tracking methods and some of the explanations are intentionally simplified, but the aim is to explain important features of the tracking landscape relevant to over 95%, if not 99%, of advertisers for an affiliate marketer to have a better appreciation of affiliate tracking.
First Party Tracking Methods
The common first-party tracking methods currently utilised are:
- A first-party cookie being set from the advertiser’s server
Both have the same end result of setting a cookie on the user’s device. We consider an advertiser with a well-implemented solution of one of the methods above to have a good first-party implementation for their browser-based users. There need to be further considerations applied regarding the tracking implementation for an advertiser that uses an app in conjunction with its website to engage with users.
There are different flavours of server tracking (sometimes called postback tracking) depending on precisely when it is initiated in the handover redirect chain and also whether it is ‘passed’ a network identifier or it ‘makes a call’ for it to be generated. The nature of the redirect path varies too, as it may avoid a hop through a network server.
Methods Without First-Party Cookies
There are of course other methods used in affiliate tracking including:
- Third-party tracking
- Deterministic/Probabilistic tracking
- Voucher code tracking
- Advertiser’s own implementations
The third party option has been discussed greatly recently due to the deprecation of third party cookies on Chrome happening in 2024. This aligns Chrome with the situation on Safari (following the 2019 introduction of ITP) and Firefox.
We often refer to it as ‘Fallback third party tracking’ as it is the tracking method that most commonly records the sale in circumstances where first party tracking is absent or compromised.
The ‘Tracking Hierarchy’
These tracking methods can be arranged – actually or conceptually – into a ‘tracking hierarchy’. Some networks may also use other components that enhance the efficacy of these methods listed; for instance utilising a tag on the publishers page. Networks don’t want to enter the same sale into their database twice so they’ll usually follow such a hierarchy for identifying, receiving and deduping communication to them of any sale from the checkout.
Therefore networks will listen for the sale confirmation signal following a hierarchy with the ‘inferior’ methods being considered fallback methods on the superior ones.
Some methods and/or implementations may identify transactions back to the click reference more reliably than others.
- The preferred first party methods are better.
- Tracking and attribution rules may also be referenced during the process of recording (or decision not to record) a transaction at the network after the communication from the checkout page, so these are clearly important to the reliability of tracking too.
- The advertiser may also have rules for determining whether a referral is recognised as warranting the affiliate code being triggered (e.g. is it in priority/ acceptably in duplicate to a prior social referral).
- The advertiser may also have rules for whether the checkout code should be triggered (for a similar business rationale).
- Legislation may have an impact too – see section on user consent.
All rules have three main considerations:
- Are they fair (it helps the program work if they are equitable);
- Are they communicated correctly and;
- Are they technically implemented in accord with each party’s intentions.
GA4, Adobe, Omniture and Affiliate Tracking
The implications of measurement of the same activity by two measurement approaches (affiliate tracking and analytics) are important to understand. In such situations, all of the above rules and considerations need to be reflected in the analytics handling for such analytics and affiliate measurement to align. In the case of GA4 there are other limitations that need to be considered that others have written about recently.
For fair comparisons, the analytics tagging of the other advertising channels also needs to follow how it’s intended to work, and as importantly, those utilizing this reporting need to understand how it is set up and how it actually measures. A strongly configured affiliate program is likely to be a better ‘source of truth’ for affiliate channel performance than a compromised company-wide implementation. It is completely understandable that an advertiser would like to review all its online marketing in one place.
Advertisers need to have a clear understanding of how consent is desired to be controlling the different measurement methods, and how consent actually controls the different measurement methods. Moonpull does see issues with consent platforms being poorly implemented and inadvertently breaking first party tracking.
Marketers will increasingly need to be aware that consent matters may create differences between measurement methods (sales tracked) and also the actual activity (sales resulting from the referrals).
“The big difference between the tracking implementations for me is down to whether or not there is an expectation that the attribution window listed for the advertiser will be respected or not.
If there is an expectation that a Publisher will get rewarded outside of the session (30 minutes of inactivity as standard) then, whatever tracking methodology is implemented by the advertiser, a cookie (or local storage) will need to be used by the Network/Advertiser to store the tracking information.
James Breckenridge, Moonpull CTO
An Example Hierarchy
Different networks and advertisers may use a slightly amended order, but this offers an example set up:
- Third Party tracking
- Voucher code tracking
- Deterministic tracking
We have omitted from this article – and thus the hierarchy above – a discussion of server to server tracking with the ANI handled server side throughout the user journey from landing page to basket. While this handling negates the need for permission under PECR or similar legislation, these implementations are rare.
More important than any hierarchy is ensuring the process is capturing all the clicks and transactions that are reasonably to be expected to be relevant to the commercial relationship between publisher and advertiser. With the deprecation of the third party method on Chrome this shines a light on the first party methods. We consider these methods to have more ‘moving parts’ and therefore require maintenance and monitoring to ensure they’re working reliably on all intended pages.
This is far from an exhaustive analysis but will give affiliate marketers more of an understanding of the processes involved in devising and providing tracking technologies. It should also help advertisers with appreciating the importance of setting up analytics platforms to correctly reflect the affiliate activities alongside other referral channels.
It is early 2024, with Chrome on the path to deprecating the availability of the reliable fallback third party tracking from the toolkit for affiliate marketers. This is shining a light on the first party methods and implementations. We consider these methods to have more ‘moving parts’ and therefore require maintenance and monitoring to ensure they’re working reliably on all intended pages. We would encourage all affiliate marketers to appreciate their implementations better to be well placed for the journey ahead and building strong relationships with their partners within the channel. We encourage talking about tracking, and we recently published a common language that may be used to may conversations accurate and efficient.