Singular's Assists feature provides multi-touch attribution (MTA) visibility, offering a complete view of the user journey from impression to final conversion. Assisted Installs quantify the role of channels that contributed to an install but did not win the final last-touch attribution. The MTA Installs metric uses customizable modeling to allocate partial and full credit to all contributing touchpoints, providing a clearer measure of overall channel performance.
By using these metrics, marketers can understand channel overlap, prove the true value of upper-funnel activities like CTV, and make more confident, data-backed budget allocation decisions. This moves beyond last-touch models to give the clarity needed to optimize spend and prove incremental impact.
Core Concepts
- Assist and Multi-Touch Visibility: Marketers are increasingly using channels like CTV and video networks. Traditional last-touch models can obscure the impact of these channels. Assists addresses this by showing touchpoints that did not win the last-touch attribution but did contribute to the install.
- Touchpoint Waterfall (T-1 to T-4): Singular tracks and reports up to four additional touchpoints leading up to the install, in addition to the Last Attributing Touchpoint (T0). These are referred to as T-1, T-2, T-3, and T-4, representing the touchpoints immediately preceding the last attributed one.
- Deduplication Logic: To prevent sources from incorrectly assisting their own installs (a "spamming" use case), Singular will dedupe per source. Singular will report both a click and an impression within the T1-T4 window only once per source.
- Attribution Logic: Whatever attribution windows have been configured per Source/app/tracking link will still apply the to the assist and MTA logic. Advertisers that prefer longer attribution windows for views and for CTV views can configure this in Partner Configuration.
Aggregated Metrics in Reports/Pivots-
The following metrics are available at every granularity in your Singular reports, offering measurable clarity on assist and multi-touch performance:
| Metric | Definition | Purpose |
| Assisted Installs | Installs where a source had a touchpoint in the T1-T4 window but did not win the last-touch attribution. | Quantifies the uplift a network provides by contributing to installs that are ultimately attributed to a different channel. |
| Single Attributed Installs | Installs attributed to a Last Touchpoint that had no other source touchpoints in the user journey. | Measures the number of installs driven solely by one network, helping assess its unique audience. |
| Co-attributed Installs | Installs attributed to a Last Touchpoint that also had at least one other source touchpoint in the journey (T1-T4). | Measures the amount of overlap and cannibalization a network experiences or contributes to. |
| MTA Installs |
The total modeled installs, calculated as the sum of Single Attributed Installs and Modeled Partial Attribution (based on configured weighting). The default formula is: MTA Installs = (SAI × 1.0) + (CAI × 0.3) + (AAIT-1 × 0.7) Where:
|
Provides a single, modeled install count that credits all contributing touchpoints according to your business rules. |
| MTA eCPI | The total cost divided by the total MTA Installs. | Calculates the true cost per install by factoring in the cost of all contributing touchpoints, moving beyond last-touch cost calculation. |
Configurable MTA Modeling
Singular offers configurable, basic Multi-Touch Attribution (MTA) modeling. This is done via a weighting formula that can be customized per organization to reflect your specific definition of channel value.
A critical rule for this modeling is that the total number of last-touch attributed installs must equal the total number of modeled installs.
- Single Attributed Install = 1 full credit
- Co-attributed Install = 0.3 credit
- Assisted Install = 0.7 credit
The MTA Installs metric will reflect the sum of these credits, providing a view of channel impact based on your business logic.
Additional User-Level Insights
User-level data for Assists offers granular transparency by providing details on every contributing touchpoint—including the regular supported breakdowns (source, campaign, touchpoint type (click/view), timestamp, and attribution method) as well as:
- Assist touchpoint count (total number of Sources for a given install). This value will only be reported in touchpoint 0. See examples below for clarification.
- Assist Level - what was the sequence of touchpoints
- Install ID - this is a unique identifier that allows you to group the relevant touchpoints per install
This allows marketers to analyze the exact sequence of interactions, determine channel overlap on an individual basis, and understand the specific customer paths that lead to an install.
An example of this will look like:
User-Level Assists Data Example
The following example illustrates how an install with 4 touchpoints is recorded in the user-level data, demonstrating the use of the Install ID, Assisted Level, and Assist Touchpoints Count.
| Device ID | Device ID Type | Tracker Name | Partner | App Longname | Platform | Assisted Level | Assist Touchpoints Count | Install ID |
|---|---|---|---|---|---|---|---|---|
| 598a388d-b4f6-4318-90f5-a80e14230cf8 | gaid | test1 | com.singulartest | android | 0 | 4 | b8ac7f0b-c622-43d4-b5e2-e4144000e55d | |
| 598a388d-b4f6-4318-90f5-a80e14230cf8 | gaid | test2 | Hehemobi | com.singulartest | android | 1 | b8ac7f0b-c622-43d4-b5e2-e4144000e55d | |
| 598a388d-b4f6-4318-90f5-a80e14230cf8 | gaid | srfs0 | Unity | com.singulartest | android | 4 | b8ac7f0b-c622-43d4-b5e2-e4144000e55d | |
| 598a388d-b4f6-4318-90f5-a80e14230cf8 | gaid | testse0 | UPSMOBI | com.singulartest | android | 3 | b8ac7f0b-c622-43d4-b5e2-e4144000e55d | |
| 598a388d-b4f6-4318-90f5-a80e14230cf8 | gaid | linli | recomob | com.singulartest | android | 2 | b8ac7f0b-c622-43d4-b5e2-e4144000e55d |
In this example we can see for a given install there are 4 touchpoints. This is understood by looking at the row where Assisted Level=0 and Assist Touchpoint Count =4.
We can also understand the sequence of Partner Touchpoints for this install.
Marketers can build additional dashboards with user-level data to map out Partner Specific Overlap and Partner Sequence.
Marketers can use these metrics to evaluate their sources by moving beyond a simple last-touch model to assess a channel's true incremental value. By analyzing the volume of Assisted Installs, they can prove the impact and contribution of upper-funnel and supporting channels, like CTV or video networks, that might otherwise be overlooked. Comparing Single Attributed Installs to Co-attributed Installs reveals how unique a source's audience is and quantifies the overlap or cannibalization between different networks. Finally, they can use MTA eCPI to calculate the real cost of an install across all touchpoints, enabling more accurate ROI validation and confident, data-backed budget allocation
Comments
Please sign in to leave a comment.