An attribution model is the framework that decides how credit for a conversion is assigned across the touchpoints in a user's journey. Most installs and purchases follow more than one interaction, and the model is the rule that says which of those interactions gets the credit. That rule shapes how every channel looks in your reports.

How attribution works

When a user converts, the measurement system looks back at the clicks and views that preceded the event and applies a rule to distribute credit. The rule can hand all the credit to one touch or split it across several. Because budget follows the channels that appear to convert, the rule you pick quietly steers where money goes.

Common attribution models

• Last-click: all credit goes to the final touch before conversion. Simple, but it ignores earlier influence.

• First-click: credit goes to the first touch, useful for measuring discovery.

• Multi-touch: credit is split across several touchpoints for a fuller picture.

Last-click vs. multi-touch

Last-click is easy to run and easy to explain, which is why it stays common. It also flatters whichever channel happens to sit closest to the conversion, often retargeting or branded search. Multi-touch spreads credit across the journey and gives a fairer read of channels that drove discovery, but it needs more data and more careful setup. Many teams run both and compare.

The attribution window

Every model runs inside an attribution window, the length of time a touch can still get credit for a later conversion. A click window might be seven days and a view window one day. The window is as consequential as the model itself. Set it too short and slow conversions look organic; set it too long and unrelated touches collect credit they did not earn. Advertisers and publishers agree the window up front so both sides count the same events.

How to choose a model

Match the model to the journey. Short, single-touch paths like a rewarded install are read well by last-click. Long consideration paths, common in higher-priced purchases, need multi-touch to show the channels that opened the journey. Start from the question you are trying to answer, whether that is where to cut spend or which channel drives discovery, and pick the model that answers it honestly.

Why it matters

Your attribution model decides where budget looks effective, so it directly shapes spending decisions. The wrong model can starve a channel that is quietly driving conversions.

Attribution and offerwalls

Because offerwall and other incentivized traffic convert on completed actions, clean attribution is essential. A Mobile Measurement Partner (MMP) and server-to-server postbacks confirm exactly which offer drove which install or event, so publishers and advertisers agree on what happened. The completion itself is the attributed event, which removes much of the guesswork that clouds view-based models.

Common mistakes to avoid

• Treating one model as the truth. Each model answers a different question. Reading only last-click hides the channels that start journeys.

• Ignoring the attribution window. Too short a window drops slow conversions; too long a window credits touches that had no real effect.

• Skipping server-side confirmation. Without postbacks, completions can be miscounted or disputed.