Why Track Flash Interactions?

At least three product enhancements rolled out in recent weeks aim to give marketers detailed information about how Web users interact with Flash ads and apps. But do Flash interaction metrics say anything about the branding impact of an ad or Web site?

DoubleClick’s Motif rich media solution can now report in depth on how a user behaves within Flash ads, NetIQ has rolled out a WebTrends toolkit letting marketers track interactions within Flash elements, and the latest version of AtlasDMT’s ad management platform includes a brand exposure duration (BXD) metric that claims to quantify the value of online rich media.

While the growth of rich media is undoubtedly fueling the flurry of new product features, measuring interactions is nothing new. Bluestreak has long offered it, first as part of its Java-based rich media creation and delivery system and later in Flash. Bluestreak Director of Product Management Eric Picard said the company had trouble persuading marketers there was value in tracking and reporting on user interactions within Flash ads.

“What we decided based on our conversations with customers at the time was that there isn’t enough data about what ‘interaction’ means,” Picard said.

What Bluestreak specifically tracked was the amount of time an ad was expanded, since its technology, by default, enlarged every ad that was clicked on. However, Picard said, “Even with that, we found it difficult to convince advertisers there was a correlation to branding.”

Now a new wave of interaction metrics is cresting. With its formal release this month, Motif reports on approximately 30 unique interactions, including ad exposure time, interaction time, whether an ad was fully delivered, how much clicking was done, when and for how long a movie was played and so on. Scott Spencer, product director of DoubleClick’s DART for Publishers group, characterized Motif’s innovation over competitor products in terms of the control it gives to the creators of ads.

“Motif provides the ability for designers to decide what they want to track,” he said. “There are certain standard metrics, such as exposure duration, which of course we’re providing. But the real power of the product is the flexibility to track whatever you want.

“It lets you figure out how many people moved their mouse over a particular pixel,” Spencer offered by way of example.

But that still leaves the question: Are marketers more likely now to perceive value in these metrics than they were before, when Bluestreak started pushing them?

“We’re not making any claims yet,” Spencer said. “My hypothesis is that there will be some significant correlations [between branding impact and interactions]. But there’s certainly no guarantee… I think what we’re talking about is a tie that needs to be made. There’s not enough empirical evidence out there.”

He said while DoubleClick had not yet done so, the company hoped to offer valid statistical support linking brand impact with various interaction metrics, including exposure duration and type of behavior. (He also said doing so wasn’t DoubleClick’s top priority.)

In the meantime, he suggested advertisers and agencies would find ways to make the metrics useful, and said the value of interaction reporting will always depend on the campaign.

“The value agencies provide is analysis of what’s important to look at,” Spencer said. “I don’t think what our ad interaction metrics show will have the same relationship to brand awareness for a deferred action campaign as for a direct response campaign.”

Atlas DMT, which is pushing its own brand exposure duration metric for rich media, acknowledged no statistical correlation yet exists between interaction and impact. However, Young-Bean Song, the company’s director of analytics, argues a logical tie exists.

“It’s not too big of a leap to believe the longer somebody is exposed to an ad, the greater the brand value of that exposure,” he said.

Young hopes marketers will embrace the exposure duration metric as a complementary third dimension to their reach and frequency reporting.

“Imagine two campaigns. One delivers 20 hours of brand exposure time and another delivers 100. I’d naturally go with the more-is-better scenario,” he said.

Nate Elliott, an associate analyst with Jupiter Research, disagrees that duration can be used as a direct proxy for brand impact. Elliott agrees length of exposure likely has some branding effect, but he said other factors such as creative quality and content targeting are more important. And he reiterates none of the companies offering interaction metrics has offered a statistical connection. Jupiter Research shares a parent company with this site.

Karen Anderson, media director for interactive agency Modem Media, is also reserving judgment.

“I can’t say whether there’s a definitive relationship between these interaction metrics and branding,” she said. “To find out, you’d have to combine behavioral data with attitudinal data.”

But Anderson clearly suspects a branding connection, and she said regardless of whether one is made, Modem Media can make valuable use of behavioral reporting such as that offered by Motif — which the agency is now testing.

“The value is in seeing what a person is interested in and how they interact with the brand,” she said. “We can use that to build better media plans and smarter and better creative.”

Which is exactly how most media planners will likely use the duration and other interaction metrics, at least until the companies reporting them come up with a statistically valid branding link.

“The question becomes, how do I take all this data and figure out how I’m going to use that in an actionable manner,” said Bluestreak’s Picard. “If you can convince your customer that interaction time is a corollary to branding, that’s great for you. If I were a media planner at an agency, I’d be reluctant at first to trust this.”

Picard also hinted the seemingly sudden appearance of these metrics in everyone’s ad serving platforms may be driven less by innovation than by competitive pressures.

“My sense is this flurry has happened for competitive reasons,” said Picard. “The landscape of ad serving is very competitive from a feature-to-feature standpoint. The RFPs have checkboxes, and if you don’t measure up to competitors, you won’t be called in for a meeting.”

Even if a correlation between branding and behavior is eventually made, marketers should be careful about giving too much attention to the branding significance of interaction, since branding can and does happen without it. Dynamic Logic has built its business measuring the brand impact of ads that are never clicked on, clicked through or otherwise interacted with.

“Only a small percentage of people will interact, no matter what the creative,” said Michael Spitalnik, a regional sales manager for the measurement firm. “Five percent would be a high number. That still leaves 95 percent who are not. But they are still interacting attitudinally.”

That’s not to say measuring interactions within Flash ads is a mistake, Spitalnik said.

“The mistake,” he said, “would be to only measure interactions.”

News Around the Web