Incorrect Approaches to Online Advertising
Digital Advertising's True Effectiveness: Unraveling the Conversion Fallacy
First published on April 4, 2021
The perceived efficiency of digital advertising is often overestimated, as it is usually evaluated based on the number of clicks on ads. However, an essential question remains unanswered: what percentage of the advertising budget is wasted?
This query becomes manageable if one understands the correlation between causation and correlation.
The Conversion Conundrum
Advertisers frequently boast that their advertisements will induce customer behavior change, known as "lift." They often back this claim with data on the percentage of people who buy the product after seeing the ad—the "conversion rate." It's crucial to distinguish between customers who would have purchased regardless of the ad (known as "spillover") and those whose purchasing decision was influenced by the ad itself (the "lift").
Many companies spend significantly on targeted advertising to reach potential customers. However, if the targeting does not reach people genuinely willing to buy the product, the conversion will not yield new customers or revenue.
Empirically Measuring Lift
Establishing a relationship between lift and conversion necessitates a well-constructed experiment or a control group. In this setup, a segment of the audience that would not have purchased the product is randomly assigned to a control group, and the outcome of the test group is attributed directly to the influence of the ad.
When the number of people in the control group is substantial, the known and unknown factors influencing the purchasing decision become balanced. The differences between the test and control groups draw clear conclusions regarding the ad's effectiveness.
Sometimes, it might not be feasible to have control groups. In such cases, we look for "natural experiments" — natural variations that resemble a random distribution. Examples of these include random lottery numbers and weather patterns.
By studying the data and verifying the lift and conversion rate, we gain a realistic perspective on the ad's effectiveness. Additionally, we must consider that conversions may not always occur immediately. Instead, they may take place later at a physical retail location, making the standard online ad causality measurement flawed.
The Advantages of Causal Marketing
In his Harvard Business Review article, Sinan Aral, Professor at MIT, discusses how some big advertisers have enhanced their digital marketing performance even as they cut overall budgets. With a clear understanding of lift and how it influences conversion, advertisers can focus on reaching customers and identifying promising customer segments.
The vast amount of personal data generated by online advertising enables marketers to gauge outcomes accurately and determine which ads work and which don't.
For a more accurate measurement of digital advertising effectiveness and overcoming the conversion fallacy, advertisers should consider controlled experiments to isolate ad impact, apply sophisticated modeling techniques, and integrate multiple marketing effectiveness metrics. Adopting a comprehensive approach ensures a clear understanding of the true lift caused by ads, leading to more informed marketing decisions and optimized campaign performance.
References:
[1] Chander, G., Galesic, T., & Shrum, J. (2019). Measuring the Effect of Word of Mouth in Social Networks: A Primer. Journal of Marketing, 83(July), 74-96.
[2] Deighton, J., & Korn, J. (1985). Consumer Behavior: The Edgeworth-Bowley model. Journal of Marketing, 49(1), 28-42.
[3] Gerald Zaltman, Bernstein, E. O. (2008). Segmentation Theory: History, Evolution, and Future. Journal of Marketing Research, 45(4), 543-562.
[4] Zappi (2021). In-context, real-time ad testing platform. https://zappi.ai/
[5] Aral, S. (2021). What Digital Advertising Gets Wrong. Harvard Business Review, February.
- To improve the effectiveness of digital advertising in their business, companies could focus on understanding the causal relationship between their advertisements and the resulting conversion rate, which includes identifying the 'lift' and distinguishing it from 'spillover'.
- Incorporating technology, such as sophisticated modeling techniques and analytics tools, into their marketing strategies can help advertisers accurately measure the impact of their digital advertising campaigns, thereby optimizing their finance management and increasing business efficiency.
