Using attribution analysis to drive efficiency in your campaigns

The continued growth of digital advertising has diversified the range of opportunities for marketers. This evolution has also surfaced a daunting set of challenges, particularly when it comes to attribution. Which formats or platforms—and what level of investment—produce the best results?

Marketers often use assumption-based approaches to understand the impact of their media investments. Unfortunately, simplistic methodologies like last-touch, last-click, and others can rely on rules and presumptions, rather than data and science.

As a result, conclusions can be inaccurate, leading to incorrect conclusions about which investments are working and which aren’t. Many marketers understand that these problems exist but believe they have few options at their disposal.

Understanding the value of bought media

At Yahoo we’re able to provide our clients with a proprietary, patented approach to attribution that incorporates two key touchstones we believe are vital for accurate advertising measurement:

  • Multi-Touch: Our model distributes credit using an algorithmic approach that takes into account the impact of each impression from each channel.

  • Incremental: We employ an exposed versus control framework to measure the incremental impact of each placement.

This approach is unique because it not only considers and evaluates all impressions that occur before a conversion takes place, it also considers the effectiveness of the ads themselves. The result for marketers is a more accurate view of campaigns, which leads to greater success and return on investment.

Benefitting financial advertisers

We’ve seen this methodology prove especially beneficial for financial advertisers, where competition is fierce, margins are tight, and every ad dollar matters.

We recently worked with one of our advertisers, a large financial services company, to drill into its attribution analytics. We measured results across 11 different placement groupings, categories including managed performance targeting, native, branded and non-branded search, log-in pages, and more.

All but one category showed lifts in site conversion as compared to a control group, but when it came to total incremental conversions, just four categories racked up nearly 80% of the total number: native, two varieties of managed performance targeting placements, and our “Travel Splash” page. But that only tells part of the story.

Next, we looked at cost-per-conversion, and discovered wildly disparate results: The two managed performance targeting placements had cost-per-conversion rates that were more than 90% lower than the native category, and a fraction of a percent of the cost for conversions from the Travel Splash page.

Clearly, these two placements performed the best in both driving incremental conversions, and in efficiency. This insight led Yahoo to recommend a reallocation of ad spend for our client, which, when modeled out, would lead to:

  • No increase in budget
  • 70% increase in impressions
  • 53% increase in number of converters
  • 34% decrease in cost-per-conversion

As a result of this analysis, we were able to uncover strategies for much higher campaign efficiency. With the fragmentation in our digital world, marketers understand the need for attribution in their campaigns.

But rather than treating attribution like a glorified accounting exercise, they should think of it from the perspective of ad effectiveness. That is, measure the causal relationship between ad exposure and intended outcomes, then use all exposure interactions to give credit where credit is due.