Video Is Next For Chartbeat's Real-Time Engagement Analytics

chartbeatChartbeat recently announced that it has brought its "engaged time" metric to its growing web ad analytics offering.

Chartbeat CEO Tony Haile and Product Manager Alex Carusillo discussed their company's objectives in advertising and industry trends with AdExchanger...

AdExchanger: Overall, where does the analysis of advertising fit in Chartbeat’s strategy?

TONY HAILE: Our core has always been helping people align their teams with their metrics to get the most out of them and adapt in real-time parameters. With advertising, we’ve found a recurring misalignment between the value of an ad and how much we’re paying for it. We’re bringing those things into alignment now because it touches on whatever we do around content and engagement. It’s certainly important to Chartbeat.

Given your unique position in analytics, in your opinion, is it possible to achieve cohesive insights about content and ads – and commerce?

TONY HAILE: We're seeing people move from having only one way of generating revenue – as in “I’m just going to have some ads on my site” – to having multiple ways of generating revenue, whether that’s commerce, through subscription services, or a range of other things. It’s not just about indiscriminately chasing traffic anymore; it’s about building a loyal, returning audience and understanding which audience members are great candidates for which path such as subscription, commerce and who is an attractive audience for advertisers. And then the challenge is making sure that the right messages and paths are set in front of these people. We believe the best indicator is the amount of, what we call, “engaged time” that you're able to capture - in other words, the amount of audience attention you're able to hold.

Unsurprisingly, the more engaged time you capture, the more likely someone’s able to return, and you can monetize them with more than just the classic eyeball flash.

From an advertiser’s perspective, what makes Chartbeat’s engagement metrics different than others in your competitive set?

TONY HAILE:   You have a variety of workable things in the context of the ad, and they’ve all tended to be fairly explicit actions. Whether it’s a mouseover or a click or a share or whatever – those are new kinds of proxies for a click-through rate. Because we have our code on the page as well as in the ad, we can tell not only whether the ad is in a visible place; we can tell if that’s actually an active window, if someone moved their mouse and scrolled at all in the last few seconds to suggest they’re really there. And we implement a whole range of different metrics: Is this person actually there, looking at the content in front of them, or did it load in the top 700 pixels of the page, or has their mouse scrolled over something?

There is a linear correlation between engaged time and brand recall. So it’s not just about viewability. It’s about being in an engaged state with that ad in view for x number of seconds. That matters to brands because then they can understand the qualitative difference between an impression over here – which saw maybe a couple seconds of engaged time – and an ad in another spot for the same audience that had 20 seconds of engaged time and is far likelier to get brand recall.

So in terms of brand recall analysis, what is the difference between, say, a Dynamic Logic study and Chartbeat’s offering?

ALEX CARUSILLO:  I don’t actually consider the two as conflicting in any way.  Dynamic Logic asks people how the viewer’s feelings have changed based upon exposure; what we’re talking about is what happens during that exposure. Dynamic Logic says whether they saw or didn't see an ad, and whether it happened 5, 10 or 15 times. And that’s great - that’s how you quantify the change.

We believe that what a company like Dynamic Logic has been looking at – exclusively the exposure – isn’t enough. It’s got to be people’s actual interaction with those things that are in view, because seeing an ad for half a second still counts as seeing an ad for half a second. Seeing an ad for three seconds – that’s different. We’re here to illuminate what Dynamic Logic is missing – what actually happens when those ads are fired.

This data would be very valuable to the marketer doing the buy on the website. Are there APIs that could filter back through the marketer’s ad server and provide some ability to optimize? How can Chartbeat address the buy side?

TONY HAILE:  What we’re doing even now is helpful to both the publisher side and the buy side because both sides better understand the true value of the ad they're putting in place, discovering opportunities for a brand to more effectively put their budget to work with the most engaging ads. Both sides win to a large extent. We’re not trying to give publishers an edge over brands; we see if the effectiveness of your ads has been murky, and then we bring clarity. Over time, we want to make it easier for brands to understand this directly with tools they can use themselves. For the moment, we’re focused on just making sure we bring absolute clarity of data to the people.

It would seem you have to deal with the conundrum of potentially serving the sell and buy sides simultaneously. True?

TONY HAILE:   That’s indicative of the challenges we’ve had measuring value in this industry as a whole. Both sides want audiences to be engaged. They want their brands to be associated with great content, and what we’re enabling here is an alignment between two groups that have traditionally been misaligned. We care about being able to say, "Can we bring these people together around some real metrics versus the kinds of proxies we’ve been using thus far?"

What about mobile advertising? Are there unique engagement metrics or other challenges there?

TONY HAILE:   Whether it’s mobile, tablet or desktop, it’s all the same to us. No matter what, we pull in that attention data, that engaged time, so the source of their attention doesn’t matter to us as much; it’s whether we’re actually capturing that attention. So mobile works for us just as well as anything else.

And video?

TONY HAILE:  We’re actually releasing a trial for video in the next month or so, and we’ve been doing a tremendous amount of work around that. We’re going to start providing tools for editorial - for producers and so on. Traditionally, video has been kept utterly apart from the rest of the team within publishers. There hasn’t really been a holistic view of how video integrates with the rest of the site and how to drive traffic and engagement between the two. So we’re going to be putting that out in April.

One of the great things about these types of metrics is the relation between video and TV, in general. You have a way for brands to understand that they can effectively buy 2 million engaged minutes in TV and then do the same thing on the Internet.

What do you see as maybe a key analytics issue that will be coming together or addressed in the future?

TONY HAILE:   Rather than jumping on something like attribution, I think there’s going to be a few general trends over time. As our environment becomes faster and more complex, you see a much more rapid pace of adaptation across the space, and not just in terms of real-time bidding but everywhere. From an analytics perspective, that means more and more people within the organization are going to be using analytics. It’s not just for your data science team; it’s going to be your frontline team doing this, and data won't just be a destination where you pull a report at the end of the week; it’s going to be an ambient part of your environment.

If you want an analogy, the world of analytics used to be an aircraft’s black box that you analyze once the plane has crashed. Now, analytics means truly being in the cockpit, seeing the data as it happens, being able to make good, strong decisions supported by technology on that front, and then acting to see feedback take place. Over the next few years we’ll see that move from black box to cockpit.

Follow Tony Haile (@arctictony), Charbeat (@chartbeat) and AdExchanger (@adexchanger) on Twitter.

 

 

Add a comment

XHTML: You can use these tags: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>