Capture the effectiveness of your TV / Radio campaigns in Billy Grace. Learn more in this article about how the results are modelled.
Modeling & Interpretation
TV and Radio are added as separate clusters within the Unified Marketing Measurement (UMM) session models. This means:
Impressions or GRPs are not used directly in the model. Instead, we apply adstock and saturation transformations to account for delayed effects from views to actions and for diminishing returns;
We allow for time-varying effectiveness. For instance, TV campaigns with promotional messaging during the Black Friday period often have a very different impact on sessions and conversions compared to brand-building commercials aired during regular periods. Typically, we observe a higher ROAS for TV during intense promotional periods, suggesting that customers are more responsive due to discounts;
Our UMM approach also accounts for trends, seasonality, outliers, non-media effects, and other external factors. This helps prevent over-attribution to media, particularly during high-activity periods like promotions, where sessions and conversions may increase regardless of media spend.
It is important to note that our model measures the view impact on sessions, and links these views to conversions through MTA-based attributions of the sessions. As such, the approach captures short- to medium-term impact, and does not reflect long-term benefits typically associated with consistent brand-building efforts via TV or Radio.
Therefore, we recommend using our attribution model as part of a broader evaluation strategy that includes the effect of TV and Radio on brand KPIs and overall brand-building performance.
Data Quality
We only integrate TV and Radio data into the UMM session models if there are at least 20 days of available data, as is standard for all digital media channels.
This 20-day requirement refers to the total number of data points per channel, not per flight.
The data can be measured in either impressions or GRPs. Since we model different media channels as separate clusters, a consistent cross-channel or cross-platform definition of views is not required. However, consistency within the same channel is essential.
For example, the GRP definition for all TV campaigns must be consistent and based on the same target audience. The GRP definitions for TV and Radio may differ, as they can be measured based on different target audiences.
Glossary
Adstock
Adstock is a marketing concept that models how the effect of advertising decays over time, accounting for the lingering influence of past ad exposures on consumer behavior.
Diminishing returns
Diminishing returns in marketing refers to the principle that as you increase investment in advertising, each additional unit of spend generates progressively smaller gains in outcomes like sales or awareness.
GRP (Gross rating point)
Gross Rating Point (GRP) is a measure of advertising impact, calculated by multiplying the percentage of a target audience reached by an ad (reach) by the frequency of exposure, indicating the total impressions relative to the audience size.