The Be-All and End-All Metric for Evaluating Online Ad Performance

Performance measurement in the digital display ad industry is flawed. Traditional metrics like impressions and clickthroughs leave advertisers paying for quantity without weighing the quality of the influence ads have on consumer shopping behavior.

There are many granular problems associated with traditional metrics, but the underlying reason impressions and clickthroughs don’t work is their failure to effectively align campaign goals with broader marketing goals. While marketing objectives focus on increasing conversions and building brand loyalty, campaigns optimized around traditional metrics are designed to maximize only the specific metric for which they’re accountable. Here’s how this can be detrimental to the success of an online ad campaign:

Impressions
If you’re optimizing a campaign around impressions, the goal quickly shifts from impacting consumers to delivering the highest volume of ads possible within your budget. Theoretically, more ads means more brand exposure and, ultimately, more conversions. But in order to maximize the amount of budget that can be spent on increasing impressions, advertising partners often dump ads into cheap inventory — e.g., below the fold or layered beneath other ads — where they're unlikely to be seen by consumers. With limited consumer visibility, it’s unlikely that an ad campaign run on impressions will make much of an impact on consumers.

Viewthroughs
A viewthrough, also called a post-impression conversion, is when someone is served an ad and takes no immediate action but later returns to the advertiser’s site and converts. Like impressions, viewthroughs leave little accountability to prove the ad was actually viewed rather than just served, creating many of the same misalignments associated with impressions. Recent RealVu data shows that up to 90 percent of ad exchange inventory isn't viewable, making it difficult to defend claiming credit for every purchaser to whom an ad was served.

Clicks
Clicks provide a simple solution to the call for a metric that proves a consumer saw an ad. However, optimizing a campaign around clicks alone oversimplifies things. Clicks prove that consumers saw campaign ads, but they don't prove that the ads influenced shopping behavior. According to comScore, just 16 percent of internet users actually click on ads. Half of those clicks account for 85 percent of the total number of clicks on display ads. This means that in order to optimize the number of clicks to report in campaign analytics, the most effective approach is targeting the 8 percent of internet users who are most likely to click ads, neglecting more than 90 percent of potential shoppers in the process.