Putting a brand on Web metrics
Comscore's Gian Fulgoni on why a click is not enough and what brand marketers should do about it
Q: What is the current metric for online ad effectiveness?
A: A third of marketers are still using the click to evaluate campaigns. Using the click as a sole metric is a problem. We had such high, but maybe naïve expectations of how advertising should work online, and it led us down the wrong path. You can use immediate metrics and look at what happened if it's a direct response campaign, but if you want to measure the holistic effect of campaign, you have to look at cumulative impressions both online and offline. There's a lot of pressure [on marketers] to get immediate response.
Q: Do you see the Internet as a branding medium or a direct response medium?
A: I think it is both. But I think the industry has really given short shrift to branding. If you measure display in terms of what happens after cumulative impressions have built up over time among people who saw ads vs. people who didn't, we confirmed display campaigns can increase visitation to advertisers' brand sites, the number of trademark search queries, and both online and offline sales. It's pretty clear the click metric is not an accurate indicator that the ad campaign has had an effect.
Q: What is your idea of how marketers should split their online spend between brand and direct marketing?
A: We haven't been looking at Internet in the right way. We hold it to an unrealistic standard when it comes to branding. The industry needs to focus more on the branding impact of the Internet and measure it appropriately. The numbers are compelling in that regard. If people spend 25% of time on the Internet, why doesn't it get 25% of the budget? Let's divide advertising into direct response and branding. The Internet has captured 30% of all direct response ad dollars spent across all media, but it's only capturing 5% of branding dollars.
Q: Can direct's metrics help marketers measure brand impact?
A: If you apply direct response measurement metrics to a branding campaign, don't be surprised if you don't see results. To some extent it seems Internet advertising folks have ignored all the work we've done in traditional advertising. We need to make sure we know what it is we're measuring and use the right tools for the right task.
Q: How do marketers measure purchase influence across their media mix? Does search end up with most conversions?
A: Yes. Why do we attribute 100% of the sale to the last click? Attribution models say by the time the last click happens, 94% of advertising impressions have occurred. If you use a panel, you see all of the exposure and analyze how much value to attribute to each component, rather than assigning it to search. I think people are beginning to understand this. We just have to keep pushing and educating.