Making Video Personal

Online marketers have made Herculean efforts to better understand the distinct and unique needs and preferences of ever more tightly specified customer segments. In theory the goal has been not only to micro-target media but to personalize messages. Until recently, online video advertisers, however tightly focused their audience, have, by default, been stuck with delivering a one-size-fits-all creative message. Below Danny Gruber, CEO of Qmecom, discusses how the personalization revolution and one to one marketing methodology can be brought to bear in the video realm.

Behavioral Insider: What was the genesis of your approach to video personalization?

Danny Gruber:
When we began we wanted to create a platform to generate customized ads for TV and cable. What we found about 18 months ago was that there was strong demand to move into the online space where there was rich and deep data, and a higher demand for more content versions. So we moved to Flash-rich media & video.

BI: Who have been the earliest adopters of this technology platform?

The biggest challenge we've encountered is that many creative agencies have a very engrained notion about how brand ads are created and delivered. They just don't get, or better, can't easily get comfortable with the idea of customizable rich media creative.
Ironically, loyalty-based direct marketing and email marketing clients are used to using data about people to target customized offers. What we're doing is extending that discipline into video.

What they see -- and what makes what we're doing such a big departure -- is that at this point so much data is being generated through behavioral targeting that advertisers have enormous power to segment customers pretty much along any kind of criteria they want to. But at the moment of truth, as in, at the point of delivering an ad, what they're delivering is essentially the ‘same ad.'

Now we see the logical next step is creating versions of video content customized to the interests, needs, and preferences of specific target segments. We didn't want to place a cost and time barrier in creating 10, 100 or hundreds of thousands of targeted ads.
BI: What kinds of data are best suited to targeting video ads?
We often see that's it's not necessarily the most complex data sets that get you the best ‘engagement' results. Within the auto vertical, for instance, if you overlay traditional fields like ‘make and model' and then add location overlaid with demographic information like age and gender, you get a really great slice of information to custom target and create video elements from a more generic auto ad.

We see something similar in entertainment. You can take a simple offering and customize it according to the genre a consumer is most interested in, the kind of music they like, other movies they've liked, types of actors and actresses they admire.
BI: Can you cite an example of how a DR advertiser is deploying this kind of targeting?

There's a tremendous mostly still untapped crossover leverage between CRM and behavioral data. A good example is a campaign we did with a large creative agency. They were working with a travel company which was developing a loyalty program. The travel company went to the retailer and said if we use customer relationship data along with behavioral data about Web browsing and shopping, we can customize offers to better target the kind of products customers would want to redeem. Rather than just a generic guess, the targeted offers drew 150% higher response.

BI: Do you see relevance to branding in this targeting platform?

Going forward we see enormous potential for brands to maintain a media buy while being able to dramatically increase engagement with the overall campaign by specific audience segments. Rather than just repurpose a single TV ad in a generic way, they can use the original creative in a wide variety of different contexts.
Another benefit of overlaying CRM and behavioral data is to add a new dimension to targeting social media. People are tantalized by how rich the profile data of social network members is. A problem though is this data tends to get stale.  I mean there's seldom much incentive to continually update your profile. But if you can relate profiles to links around friends, communities, applications and interest groups, you can build powerful psychographic segmentation. We see the demand for larger numbers of targeted creative executions in this space to grow rapidly this year.
BI: What do you see as the focal points for innovation in the video targeting arena for the rest of this year?

Among the big challenges for the rest of 2008 are to provide a really simple way to standardize all the behavioral data being generated on ad networks. There are over 300 ad networks by our count offering BT segmentation and of course each network has a different taxonomy which makes it difficult to really leverage that data to customize video by data set. We're working on a kind of master key or translation engine which allows all the various data to be unified in simple vertical channel data sets. This will make is much simpler to for the creative and media agencies to manage campaigns across multiple ad networks.
A related project is to granularize reporting. Right now reporting usually proceeds on a top-down basis. That is, know how your campaigned performed against data segmentation. But what we envision is a more bottom-up reporting, where you begin to get reports on micro-segments, by creative element and by personalized offers in real-time.




Next story loading loading..