Why Affiliate Tracking and Click Attribution Fall Short in Programmatic Advertising

Affiliate tracking and click attribution are not broken. They are precise instruments built for a narrower job than the one operators now ask them to do, and in programmatic they are asked to measure a journey that mostly happens where they cannot look.

This is what each one was designed for, where it stops, and what the gap between a click-only report and a measured one actually looks like when someone bothers to check.

What affiliate tracking was built for

Affiliate tracking exists to credit a specific partner for a specific referral. A publisher sends a visitor, the visitor converts, the publisher is paid. Within that transaction it is exact, auditable and entirely fit for purpose.

Programmatic does not look like that transaction. A campaign runs across display, connected TV, mobile, audio, native and digital out-of-home, and the same person is reached several times on more than one device before anything happens. An affiliate framework asked to describe that will credit whichever tagged link happened to be last and record the rest as nothing.

The consequence for an operator is not a rounding error. Channels that opened the sequence are systematically undervalued, and budget moves away from them — not because they underperformed, but because the measurement system had no field in which to record what they did.

Where click attribution stops

Click attribution credits the ad a person clicked immediately before converting. It is easy to audit and it is the default in almost every analytics tool. If someone clicks a display ad and deposits ten minutes later, it captures that cleanly.

Its limit is not accuracy. It is scope. Click attribution can only see people who clicked, which makes it a complete record of one behaviour and a blank space everywhere else. In a category where the decision to open an account is rarely made in the session that starts it, that blank space is most of the customer.

Why Google Analytics does not fill the gap

Google Analytics is a site analytics tool. It records what happens on your property and attributes it to the last click that brought someone there. That is a scope decision built into the product, not a defect — but it means a person exposed on connected TV on Sunday, who searches your brand on Wednesday and deposits on Thursday, arrives in your reporting as organic or direct.

GA4's event-based model improved cross-channel reporting. It did not move that boundary, because moving it would require observing an impression that never touched your site. The full argument is set out in GA4 vs programmatic attribution.

What the gap actually measures

Across six independent measurements — five operators, two continents, six years — the post-view share of conversions landed between 89% and 96.2%. Every one was customer-ID-level or platform-verified. None was modelled. An operator reading clicks alone is seeing roughly one conversion in ten.

The shape underneath that is a ratio: 27.8 impressions per converting journey against 0.12 clicks. Clicks are not the journey. They are a thin and unrepresentative sample of it.

One thing this is not: an incrementality claim. It says nothing about what would have happened with no media at all. It is a statement about measurement completeness — how much of a delivered outcome a click-only report can physically observe. Those are different questions and they need different tests. The evidence and the method are in click and view attribution in programmatic advertising.

The DSP is the most reliable source — and it still needs checking

In a programmatic environment the DSP serving the creative has the most direct visibility into delivery: which impressions were served, to which devices, in which markets, and which of those devices converted afterwards. That makes it the best available source for what programmatic did.

It is not an unbiased one, and no honest partner will tell you otherwise. A platform reporting on its own contribution has an interest in the result. The answer is not to distrust the number; it is to reconcile it. Platform-attributed first deposits compared line by line against your backend-confirmed customers, monthly, in writing, with every gap explained — event definitions, attribution windows, time zones, currencies, deduplication, naming and UTM logic.

Until that reconciliation exists, the honest label for a platform figure is platform-attributed, and it should carry that label in every report you receive.

What it costs when nobody reconciles

This is not hypothetical. The 2% Problem follows six sportsbook operators through 122,418 registrations and $541,548 of media. Every account ended — and not one of them ended on performance. Each operator's own systems showed them a fraction of what had been delivered, nobody reconciled the two views, and the relationship was decided by the disagreement.

The operators are anonymised and no client is named. The pattern is the point.

What honest attribution looks like

• One number governs, agreed before launch — platform-attributed or backend-confirmed — and written down.

• The attribution window is fixed per objective in the contract, and changing it requires your signature. Window-shifting after results are visible is the oldest trick in performance reporting.

• Last-touch and multi-touch are shown side by side, with the disagreement explained rather than hidden. The gap between them is where budget decisions live.

• Raw path-to-conversion exports are available on request. A desk that cannot produce full journey data is not analysing it either.

• First-party data does the work third-party signals no longer can — but only once it is reconciled, not merely collected.

The goal is not perfect attribution. There is no such thing. The goal is attribution honest enough that a budget decision made on it is defensible six months later — and that is achievable with instruments that already exist.

Testing this on a partner

If you would rather test a desk than take any of it on trust, the 20-question partner scorecard runs ThumbAd's published due-diligence standard against any partner, including us, and scores it against a bar of 24 out of 30. It runs in your browser, asks for no email, and nothing leaves the page. Our own answers to all twenty are published in full and ungated in How ThumbAd Answers the 20 Questions.

And if you want to see what measurement discipline produces rather than what it prevents, what an acquisition engine is worth covers twenty-one months of one operator's first-deposit economics, measured on deposits rather than registrations from day one.

ThumbAd is an independent programmatic trading desk in Oslo, trading on The Trade Desk across Europe, Africa, North America and Latin America since 2012, with betting and gaming at the core.