Visible World Sheds Light On Targeted Ad Effectiveness, Study Based On Billions Could Boost Perception Of TV

In a week that has begun with some troubling news for the value of traditional media outlets, an upbeat story is being released for at least one of them -- television. While newer digital media outlets have some observers writing old school media off, television may be poised to extend its perception as the most effective ad medium, according to a first-of-its-kind analysis of billions of anonymous individual TV viewing records.

Actually, the study -- which was commissioned by targeted TV advertising developer Visible World -- matched 10 billion individual TV viewing records over an 18-month period, and found that the ones that included ads targeted at those individuals generated conversion rates that were 70% higher than overall campaign results.

By conversion, the researchers mean consumers exposed to those TV ads actually went out and purchased the product or service being advertised. The reason the study is so encouraging for the TV industry is that it’s poised to move from relatively crude targeting methods based on broad demographic (age, sex, household income, etc.) and geographic data to much more finely targeted attributes that are more akin to online media targeting. Not surprisingly, the study’s sponsor, Visible World, is a leading player in those developments, and currently has deals to serve targeted TV ads with operators representing about 80% of cable TV households.

The study, which was conducted by Big Data analyst b3Intelligence, utilized the TV viewing records of one of those cable operators, whose identity was not disclosed.

“The results of this study are significant in that they are based on an enormous amount of data and a large number of campaigns,” explains Claudio Marcus, executive vice president-marketing and research at Visible World.

Marcus said the study initially was designed to help the cable operator understand the value of its own “cross-channel” promotional spots -- the ads it runs to upsell cable television subscription services to existing subscribers -- and consequently it enabled Visible World to get a clear picture on the actual conversions, tracking consumer behavior before and after being exposed to those ads. While the findings are based on cable TV subscription services, Marcus says they likely would work in similar ways for other TV advertising categories.

Because the study also knew exactly which services each of those households were subscribing to, the test could target them with 100% relevancy and efficiency -- something that might not be easily repeated in other categories -- but Marcus says the rule of thumb is the same: the more targeted the TV ads are, the greater the rates of conversion, whatever the campaigns goals are.

Because of the incredible control the study had over data about the households, individuals, products, services and ads being served, Marcus says Visible World was also able to identify other efficiencies vis a vis ad targeting -- including things like effective reach and frequency capping. Among other things, the analysts could segment the heaviest TV viewers who are most likely to see untargeted ads repeated over time versus light viewers who might see them more rarely, and adjust the weight of the ads accordingly to optimize the efficiency of the media budget.

While it may not be as easy for advertisers in other categories to gain access to that kind of explicit household data, companies like Visible World are trying to convince cable and satellite operators to bring more of it to market in an anonymized way that protects subscriber privacy, but enables more precise ad targeting.

Using the approach, he asserts, will make conventional TV planning and buying “TV on steroids.”

12 comments about "Visible World Sheds Light On Targeted Ad Effectiveness, Study Based On Billions Could Boost Perception Of TV".
Check to receive email when comments are posted.
  1. Mike Einstein from the Brothers Einstein, August 6, 2013 at 11:29 a.m.

    In an on-demand media universe, no one wants -- let alone demands -- any ads, relevant, efficient, or otherwise. We keep trying to finesse something that no one outside of the industry wants anything to do with. No wonder everyone buys Facebook. They have more of what doesn't work than anyone, and at the best price!

  2. Renny Fidlon from Audience Partners, August 7, 2013 at 10:24 a.m.

    Sorry, Mike. Studies have shown time and again that people are willing to accept advertising in return for free services. The more on-demand scales, the more it will accept advertising as well. Trust me, relevant ads are far more appealing than un-targeted ones. Don't believe me? Try shutting off your browser's cookies and see how your user experience changes.

  3. Mike Einstein from the Brothers Einstein, August 8, 2013 at 9:26 a.m.

    Sorry right back at you, Renny. "Willing to accept advertising in return for free services" is a far cry from wanting it, let alone demanding it; and a huge leap to actually watching it. And I can prove the point right now by suggesting that anyone reading this comment section has absolutely no idea whose ad graces the top of this page, despite the fact that MediaPost is a free service.

  4. Mike Einstein from the Brothers Einstein, August 8, 2013 at 10:42 a.m.

    @Renny again, let me put it this way: to suggest that I could possibly have a preference for something I don't want in the first place is an absolutely idiotic premise, and clearly not a viable branding strategy. What part of CTRs now at statistical zero don't you get?

  5. Renny Fidlon from Audience Partners, August 8, 2013 at 11:03 a.m.

    Yet when you run digital along with television brands see incredible lift along all branding metrics: http://www.thinkwithgoogle.com/insights/emea/featured/more-screens-are-better/

    You can't deny what works.

  6. Doug Garnett from Protonik, LLC, August 8, 2013 at 11:13 p.m.

    Here's the key question... How did they connect the two? Mass retailers have exceptional legal hurdles they face to even identify ethnic background of shoppers and are prevented from collecting many types of data. I don't doubt that TV is effective - it's exceptionally more effective than anything digital. But this claim is based on connecting specific people who saw ads with their retail purchase patterns and I don't think that's possible.

  7. Joe Mandese from MediaPost, August 9, 2013 at 7:25 a.m.

    @Doug, in this case, it was easy, because the retail products being converted were TV subscription services marketed to the households that were receiving the targeted ads. The operator had data on those households indicating what services they had before and after exposure to the targeted ads, so it was a complete control study.

  8. Doug Garnett from Protonik, LLC, August 9, 2013 at 12:05 p.m.

    Ah. So that means this is a highly specialized study of a purchase similar to an online purchase directly related to the broadcast medium being used. We should take tremendous care - there's no evidence to suggest that this case applies to anything broader. (This mistaken presumption is a classic error which direct marketers know quite well. The advertising process may be ideal for that sales channel - but not extend to any others.)

  9. Joe Mandese from MediaPost, August 9, 2013 at 2:21 p.m.

    @Doug, definitely. People should always take care when applying media research, including whether to correlate this directly with online purchase behavior. As we reported, it is a very specific control study that could only account for the effect of targeted TV advertising to people known to be in the market for the subscription television services being advertised. The application of those findings beyond that is up to anyone who wants to think about it.

  10. Claudio Marcus from FreeWheel, August 9, 2013 at 4:58 p.m.

    It is worth noting that when marketers with direct-to-consumer relationships execute household addressable TV ad campaigns, it is possible to do a blind match (using a trusted third party) to anonymously match the advertiser's customer records to the operator's subscriber records. It is also possible to do a similar post-campaign match to quantify the sales impact of the campaign. In our experience measuring such campaigns, it is not uncommon to get double-digit percentage lift in sales relative to control group(s) for the campaign.

    The larger point, that Joe mentions in the article, is that effective use of targeted TV ad campaigns help drive better sales results. Of course, a lot still depends on the power of the creative message. Great creative can deliver a home run but it is tough to do so on a consistent basis. Targeting the right message to the right viewer is more of a data-driven science akin to learning how to hit singles on a regular basis.

  11. Doug Garnett from Protonik, LLC, August 9, 2013 at 5:06 p.m.

    @Joe - Understand your point. But as a practitioner, it is increasingly frustrating that headlines like this drive opinion rather than considered evaluation of the strengths and weaknesses of the methodology. I think MediaPost works hard - but still find that too many headlines and articles report the researchers "claims" and so leave behind impressions about the research that aren't justified. But that's a hard editor choice. What I think's needed is research fact checking in our biz - the advertising equivalent of Politifact. To see how research like that reported here tends to play out, Byron Sharp wrote a nice post recently based on studies of medical research (which also needs a Politifact equivalent). http://byronsharp.wordpress.com/2013/08/02/new-marketing-practice-evidence-or-fashion/

  12. Doug Garnett from Protonik, LLC, August 9, 2013 at 5:08 p.m.

    @Claudio - At this point, you are making a claim but offering shallow evidence - especially if the research that can be pointed to is TV targeting of TV based subscriptions. Sometimes targeting will help. Just as many times it won't - because there simply aren't effective online or TV definitions of the target.

Next story loading loading..