Adtriba – future-proof makreting measurement & optimization

How to Measure: TV Ads

Television has been a dominant advertising medium for decades, offering businesses the opportunity to reach a wide audience with their marketing messages. However, determining the true impact of TV advertising on sales, brand awareness, and consumer behavior can be a challenging task. 

This article will delve into the various methods used to measure the short, medium, and long-term effects of TV advertising, discussing the pros and cons of peak detection, marketing mix modeling, "how did you hear about us" surveys, and brand studies.

Multi-Touch Attribution (MTA)  / Peak Detection

TV “touchpoints” can be integrated into MTA on a user level. For this, we need to run a peak detection first to identify visits that could have been driven by TV. Then, we integrate those visits into the customer journey. 

Peak detection is a method that analyzes short-term fluctuations in sales, website visits, or other response metrics to identify spikes in activity that may be correlated with TV advertising. This method helps businesses understand the immediate impact of a particular ad campaign.
Pros:

1. Easy to implement and understand
2. Can provide rapid feedback on ad performance
3. Helps identify successful ads and timeslots

Cons:

1. Limited to short-term effects
2. May not account for cumulative impact or delayed response
3. Can be prone to false positives due to external factors

Marketing Mix Modeling (MMM)

MMM is a statistical technique used to determine the effectiveness of different marketing channels, including TV advertising, in driving sales or other desired outcomes. By analyzing historical data, MMM can help businesses allocate their marketing budget more effectively and improve overall ROI. MMM is the “traditional” way to evaluate TV performance.

Pros:

1. Comprehensive approach, accounting for multiple marketing channels
2. Can be used to measure medium and long-term impact
3. Helps optimize marketing budget allocation


Cons:

1. Requires a substantial amount of historical data
2. May not capture the nuances of individual campaigns
3. Can be complex and time-consuming to implement

For MMM to work, we need rather granular data (ideally daily) on TV - ideally, spend or reach / GRP data. 

Incrementality Testing

While “true” incrementality tests, or RCTs, are not possible for TV (you cannot separate households into test and control groups and then analyze their behavior), we can still rely on quasi-experimental methods here. Synthetic control methods allow us to “simulate” the behavior of the control group in the absence of an actual one. For this, we need covariates to fit a model that simulates this control group. These covariates are usually state/country level (sales) time series for areas where the television campaign did not run. 

Pros 

1. Straightforward method to get an estimate of the “true” causal impact of a TV flight
2. Results can be used to further calibrate the other methods

Cons 

1. May be unrealistic in practice: there should not be other interfering “interventions” in the TV flight neither in the region with the TV flight nor in the “control” regions
2. The TV flight needs to have a well-defined start and end. This is not possible for the few brands fortunate enough to afford “always on” TV advertising
3. The impact estimation is only as good as the model for the synthetic control

"How Did You Hear About Us" Surveys

A simple but effective approach, "how did you hear about us" surveys involve asking customers directly about the channels through which they first learned about a brand or product. This method can provide valuable insights into the impact of TV advertising on consumer awareness and purchase decisions.

Pros:

1. Easy to implement
2. Provides direct feedback from customers
3. Can identify brand discovery channels

Cons:

1. Subject to recall bias
2. May not capture the full range of channels influencing a purchase decision
3. Limited to self-reported data

At Adtriba, we don’t provide survey services, but are always happy to try and integrate the results into our modeling should the client have them.

Brand (Lift) Studies

Brand studies are research initiatives that assess the impact of advertising on key brand metrics, such as awareness, perception, and preference. They often involve pre- and post-campaign surveys to gauge changes in consumer attitudes and behaviors resulting from TV advertising.

Pros:

1. Can provide valuable insights into the long-term impact of TV advertising
2. Helps businesses understand the influence of advertising on brand perception and preference
3. Can identify areas for improvement and future campaign optimization

Cons:

1. Can be expensive and time-consuming to conduct
2. May not provide granular data on individual campaign performance
3. Results may be influenced by factors beyond TV advertising, such as competitive activity or market trends

At Adtriba, we do not provide implementation of those studies ourselves but are always happy to find a way to integrate their results to calibrate our models further.

Conclusion

Measuring the impact of TV advertising is essential for businesses looking to optimize their marketing efforts and maximize return on investment. While each method has its pros and cons, a combination of peak detection, marketing mix modeling, "how did you hear about us" surveys and brand studies can provide a comprehensive understanding of the short, medium, and long-term effects of TV advertising. By leveraging these methods, businesses can better allocate their marketing budget, create more effective ad campaigns, and ultimately, achieve better results.

Leverage unbiased marketing measurement with Adtriba.