Singular’s MTA ROAS uplift is a change in allocated credit

Singular reports up to 50% higher Meta ROAS under multi-touch attribution than under last-touch reporting. That compares measurement models; it is not evidence that changing the dashboard created additional revenue.Index announcement.

Reconcile totals before allocating more spend

Singular’s product documentation describes configurable attribution weights and conservation of total install credit. The editorial implication is straightforward: inspect how credit moves between channels before reading a channel-level uplift as a business-level gain.

Our recommendation is to ask for the same underlying conversion population under both models. Preserve windows, eligible touchpoints, revenue horizon and deduplication rules. If any of these also changed, the comparison contains more than a redistribution of credit and needs a fuller reconciliation.

Keep the causal question open

A channel may have an important role that last touch understates. It may also receive modelled credit that differs from its incremental contribution. The model comparison alone cannot choose between those interpretations for a particular game.

Use the result to identify a channel worth examining through an appropriate experiment or other independently justified causal method. Do not automatically raise its budget by the reported percentage. The attached card preserves “up to”, the model comparison and the missing denominator so a presentation cannot turn the finding into a universal improvement claim.

Inspect the evidence

Download the source-population and comparability card. The extraction separates documented facts from our analysis and unknowns.

Further reading in the existing archive: Attribution is an allocation rule, not a truth; Your MMP and your network will never agree. These links provide background; this check does not independently verify their full contents.

Frequently asked questions

How should UA teams read Singular’s MTA ROAS uplift when it is a change in allocated credit?

Higher measured ROAS under another attribution model is not itself new cash. Singular reports up to 50% higher Meta ROAS under multi-touch attribution than under last-touch reporting — that compares measurement models; it is not evidence that changing the dashboard created additional revenue. Singular’s product documentation describes configurable attribution weights and conservation of total install credit, so inspect how credit moves between channels before reading a channel-level uplift as a business-level gain. Ask for the same underlying conversion population under both models; preserve windows, eligible touchpoints, revenue horizon and deduplication rules. If any of these also changed, the comparison contains more than a redistribution of credit and needs a fuller reconciliation. A channel may have an important role that last touch understates, or it may receive modelled credit that differs from its incremental contribution — the model comparison alone cannot choose between those interpretations for a particular game. Use the result to identify a channel worth examining through an appropriate experiment or other independently justified causal method; do not automatically raise its budget by the reported percentage. Preserve “up to”, the model comparison and the missing denominator.