Affiliate Tracking and Attribution: how to track sales, partners and commissions correctly

In affiliate programs, knowing that a sale happened is not enough. The company needs to understand where it came from, which partner took part in the journey, under which rule the result will be attributed and how that attribution will produce a commission.

That is why tracking, attribution and commission management need to be treated as related, but different, concepts.

In a simplified view:

interaction → identification → tracking → conversion → attribution → validation → commission.

When these stages are not clearly separated, problems arise: duplication, data discrepancies, incorrect commissions and difficulty auditing.

Tracking, attribution and commission are not the same thing

Tracking

Records events and identifies the origin of certain interactions.

It mainly answers:

"What happened, and where did it come from?"

Attribution

Applies a rule to determine who will receive credit for a given result.

It answers:

"To whom should this conversion be attributed?"

Commission

Applies the economic rule to the validated conversion.

It answers:

"Who should receive, how much, and under which rule?"

A mature operation needs to control the three stages in an integrated way.

How can an affiliate be identified?

Identification can occur through different mechanisms, such as:

• exclusive links;
• parameters;
• identifiers;
• promotional codes;
• landing pages;
• individual stores;
• campaigns;
• other references defined by the operation.

The chosen model depends on the commercial journey.

The central point is to ensure that there is a reliable relationship between the partner, the interaction and the eventual conversion.

What does tracking record?

Depending on the program, events such as the following may be recorded:

• click;
• visit;
• view;
• lead;
• registration;
• checkout start;
• order;
• sale;
• subscription;
• another relevant conversion.

Not every event necessarily generates a commission.

Tracking records the journey. The subsequent rules determine what has economic value for the program.

The conversion needs a reliable identity

Imagine that a consumer clicks on an affiliate link and later completes a purchase.

For the operation to work, the system needs to correctly relate:

partner → interaction → consumer or session → order → conversion.

The more complex the journey, the harder this association becomes.

Problems appear especially when there are:

• multiple devices;
• multiple channels;
• long intervals between interaction and purchase;
• promotional codes;
• simultaneous campaigns;
• different partners involved.

What is an attribution window?

A company can determine how long an interaction remains eligible to receive credit for a later conversion.

This period is normally called an attribution window.

For example, a rule may consider an interaction valid for a certain number of days.

There is no universal window suitable for every business.

It depends on the purchase cycle, product, commercial strategy and the program's rules.

First click, last click and other rules

There are different ways to decide who receives credit.

First click

Attributes the result to the partner associated with the first valid interaction.

Last click

Attributes it to the partner associated with the last interaction considered valid before the conversion.

Determining code or identifier

The company may establish that a specific code used in the purchase prevails over other interactions.

Custom rules

More sophisticated operations may use specific business criteria.

What matters is that the rule is clear, predictable and auditable.

The problem of journeys with multiple partners

Consider a situation in which:

1. the consumer discovers the product through Partner A;
2. then accesses content from Partner B;
3. later uses a code from Partner C;
4. finally makes the purchase.

Who receives the credit?

There is no universal technical answer.

The answer is a decision of the commercial model.

The platform needs to be able to apply that decision consistently.

Client-side and server-side tracking

Tracking can involve different technical layers.

Client-side approaches usually rely more on the browser and the user experience.

Server-side approaches allow certain events to be recorded directly between systems.

In modern operations, different mechanisms can coexist.

The choice depends on the architecture, the channels used, privacy needs and the rules applicable to the market.

Cookies should not be treated as the only answer

Cookies can be part of tracking, but an affiliate architecture should not depend conceptually on them alone.

There are other possible elements:

• proprietary identifiers;
• authenticated data;
• codes;
• server-side events;
• order IDs;
• campaign IDs;
• relationships persisted by the platform itself.

The more critical the operation, the more important it is to have reliable identification and reconciliation mechanisms.

Privacy and consent

Tracking and attribution deal with digital data and behavior.

For that reason, the implementation needs to consider:

• privacy;
• consent when applicable;
• data minimization;
• governance;
• market legislation;
• the policies of the platforms used.

There is no single valid configuration for every country.

The architecture needs to allow the company to apply its policies and obligations appropriately.

Cross-device

A consumer may:

1. discover an offer on their phone;
2. search again on a tablet;
3. complete the purchase on a computer.

This journey creates challenges for attribution.

Without some legitimate form of common identification, different devices may appear to be independent users.

For that reason, cross-device attribution requires adequate architecture and data, and not just an affiliate link.

Deduplication

The same conversion may reach the system through multiple sources.

For example:

• e-commerce integration;
• pixel;
• webhook;
• API;
• import;
• another event.

If there is no deduplication, the same order may be counted more than once.

A unique conversion or order identifier is often essential to avoid this problem.

Cancellations, returns and chargebacks

A conversion recorded initially may cease to be valid.

This happens in cases such as:

• cancellation;
• return;
• chargeback;
• fraud;
• invalid order.

For that reason, a commission may pass through states such as:

identified → attributed → pending → validated → eligible.

The nomenclature may vary, but governance of the cycle is important.

Attribution does not mean immediate commission

A sale attributed to Partner A does not necessarily mean the commission should be released immediately.

Before that, there may be:

• validations;
• a return period;
• anti-fraud analysis;
• eligibility rules;
• specific campaign conditions.

The architecture needs to separate attribution from commission settlement.

How to handle coupons and promotional codes?

Codes can be used as an additional attribution mechanism.

A company may establish, for example:

• the code prevails over the link;
• the link prevails over the code;
• a given campaign has a specific rule;
• the code is valid only for an authorized partner.

These rules need to be formalized in the program and reproduced by the platform.

Affiliate vs. paid media

Another challenge appears when the consumer interacts with affiliates and with the company's own media campaigns.

The organization needs to decide how these channels coexist.

The technology must provide enough data for the company to apply its rules without duplicating results.

Affiliate vs. consultant or distributor

The convergence between affiliates and Direct Selling creates even more interesting scenarios.

A sale may have:

• an affiliate responsible for the origin;
• a consultant related to the customer;
• a commercial structure;
• different compensation rules.

The platform needs to know whether:

• only one participant receives;
• both receive;
• there is priority;
• there is a split;
• certain models cannot coexist.

Again, this is a commercial decision that the technology needs to represent.

Attribution auditing

A company needs to be able to answer:

Why was this sale attributed to this partner?

Ideally, the history can demonstrate:

• the partner identifier;
• the interaction;
• the campaign;
• the timestamp;
• the rule applied;
• the conversion;
• the validations;
• the result.

Without this trail, discrepancies become difficult to investigate.

Important metrics

Tracking and attribution also produce data for management.

Some possible indicators:

• clicks;
• visitors;
• leads;
• conversions;
• conversion rate;
• revenue;
• commission;
• average ticket;
• performance per affiliate;
• performance per campaign;
• approved conversions;
• rejected conversions;
• cancellations;
• historical evolution.

The goal is not only to know who sold, but to understand the quality of the program.

What to evaluate in a tracking and attribution solution?

Tracking needs to be connected to the business

Tracking in isolation produces events.

Attribution in isolation produces a rule.

Commission in isolation produces amounts.

A truly integrated operation connects:

partner + journey + customer + order + attribution + commission + data.

This view reduces reconciliations and improves audit capability.

IDBCONNECT and performance-driven operations

IDBCONNECT makes it possible to structure different types of participants, commercial rules, commerce, commissioning, integrations and data within the same operational ecosystem.

This architecture creates a foundation for operations in which affiliates, consultants, creators and other partners can take part in commercial journeys with their own identification and compensation rules.