See the channels behind the last click
Give earlier touches a share of the credit. Compare installs and revenue with last click, see which channels assist, and follow the journeys that bring people to your app.
Talk to our team about availability for your project.
Credit for one install
Shared credit across this recorded journey.
Illustrative journey. Actual shares depend on your data.
Understand what each channel contributes
Start with the comparison. Check the assists. Then follow the journeys behind the numbers.
See where credit moves
Compare multi-touch with last click by channel or campaign. Find the sources that gain credit when earlier touches count.
Switch between installs and revenue
Keep the same total across both models

Same total. A different share.
100 installs
Find the channels that assist
A channel can help people install even when another network gets the last click. See who assisted whom, alongside the model's share of credit.
Count each assisting network once per install
Separate assisted installs from fractional credit

Google Ads closed 60 installs.
30
assisted installs
50% of Google's installs
30
assisted installs
50% of Google's installs
Follow the path to the install
Explore the recorded channel and campaign sequences. See which paths end in an install and where people drop off.
Read first assists through to the last touch
Explore journeys by network or campaign


Same total. A different share.
100 installs
Built around the journeys you can observe
A data-driven Markov model learns from recorded journeys that install and those that do not. It estimates each channel's contribution, then shares an install and its revenue across that journey. Your existing last-click attribution stays alongside it.
The model runs daily, with a 30-day lookback and a 7-day revenue window by default. Results depend on identifiable touchpoints and enough journeys across multiple networks. When the data does not meet the project's requirements, results are withheld with a reason.
Google and Meta report their own installs, so their claims are observed on converting journeys and can receive more credit than their true contribution. Multi-touch is a model of the recorded paths; proving incremental lift requires an experiment.
Frequently asked questions
The model, its coverage and how to get started.
See what last click leaves out
Walk through credit, assists and journeys with our team. Find out whether multi-touch can add useful context to your acquisition data.









































