Commentary
Cross-Platform Addressability: Q&A With Benjamin Masse
- by Charlene Weisler , Op-Ed Contributor, January 27, 2015
Here’s an excerpt from my interview with him. The full (video) version can be viewed here.
CW: Tell me about Triton Digital.
BM: Triton Digital was created eight years ago… [with] a few lines of business that include streaming, measurement, advertising, and audience management. When you listen to radio online via a web player or mobile device, the streaming of that FM station could very well be delivered by our streaming solution. To best measure listeners of digital audio from an online radio broadcast or music service like Pandora, our measurement solution analyzes each stream to make sure that each listener is an actual human, not a robot (as opposed to just measuring page views as a lot of measurement companies do today). I oversee our advertising business, and how we inject targeted replacement ads into audio streams to make them more relevant to the listener. The fourth line of business is audience management and engagement programs such as email, website creation, and rewards programs.
CW: Can this type of audio addressability be used to help target ads in local television? If so, how?
BM: We’re obsessively focused on all things audio. That said, similar technologies can be used for local targeting of streaming TV ads. The local market opportunity is huge, and there are a number of companies working to capitalize on it across all types of media.
CW: Is there a cross-platform opportunity here for an advertiser who wants to buy both local radio and television?
BM: There is certainly an opportunity here for advertisers looking to augment the power of their existing media buys. A number of recent studies have shown that audio has the power to boost the effectiveness of advertising on other channels. In fact, many advertisers currently use streaming audio to complement online and offline buys across channels. Audio – and particularly local audio – has grown dramatically in the last couple of years and has become a very attractive platform….
CW: We talk about programmatic in television but not so much in radio. What do you see as the benefit of programmatic in radio?
BM: Programmatic means a lot of things, one of the biggest being automation. We see that there are more and more requests for automation because it saves costs on both the agency and publisher side. By automation I mean that there is a brand that wants to advertise. They will hire an agency and they will ask that agency to figure out the best way to spend their advertising dollars to have brand exposure. In the old days of the “Mad Men” era, they were having lunch and talking to publishers, signing paper agreements, sending creative to the publisher and hoping that the publisher would traffic that ad the right way. We see that many things involved in automation are to simplify that process and save costs. You don’t need a lot of hires on either side to be sure that the buy is transacted and tracked correctly.
CW: So it sounds to me you’re saying that media sales may not be the best career track for someone starting out today.
BM: Media sales has definitely changed over the years, but there will always be a need for sales teams. It is similar to how investing in stocks used to require a trade executed by a trader. Today, you can log into your bank account and start trading stocks, but you still need to have advisers. With media sales becoming increasingly automated, those teams will no longer be exchanging paper or even electronic agreements. Instead, they will act as buying advisers to their clients.
I gather, then, that programmatic buying is going to be a big boon for media sellers---and buyers, I assume---by handling all of their burdensome paperwork. But what about the wonders of "audience buying" and the undreamed of improvement in targeting efficiencies?
I think audience would be better defined than the proxy of age/gender but it would depend on the segmentations used. I am curious to know others' opinions on this.
Charlene, while the ideal is to know the exact purchase patterns or product use/brand preference intentions of every member of the potential audience, this is simply not possible. Despite all of the talk about "big data" being used to provide a better way to deal with this, at the end, all that many of them are talking about is some sort of simulation---using demos like sex, age, and, I hope income, presence of child, ethnicity and geography----to estimate whether a Nielsen sourced "viewer" is a marketing prospect. In reality, a more direct match up between purchase/buying actions and TV viewing is already available from MRI, but this is naver mentioned as MRI isn't electronic and it doesn't use a sample size in the millions. I suspect that if MRI's "single source" indices of product use/purchase, etc. on a show by show basis were compared with the findings of the proposed big data/Nielsen simulations, you would probably get the same answer most of the time and, possibly, the MRI-Nielsen interface would be the better one.
Continuing my answer, it is simplistic in the extreme to assume that all an advertiser wants to know is whether a member of the audience is a product user or a brand buyer. Many ad campaigns attempt to single out mindset segments within the product user base----like those who are price conscious, those who care about health or ecological issues, those into image or status, the convenience oriented, the quality oriented, etc, and some of these segments are targeted in combination. How do you determine any of this using a lot of electronic shopping data----big sample or not? Again, MRI, which asks its respondents many mindset questions could provide an answer, using several years of data to provide a big enough sample----but this already existing service seems to be forgotten in the addressable media pitch.
A final comment. When advertisers buy media and use "targets" like men 18-49 or women 25-54, this has little to do with the way their ad campaigns are fashioned. The much cited age/sex "demos" are merely a way to create a single audience "currency" for each buy---mainly for CPM or other cost efficiency machinations. The reason for this is the assumption that this is all that a media buyer can do, when, in reality, media buys could be organized in a manner that was much closer to the specific targeting goals of the ad campaigns----if only the media and creative functions were fully integrated. As they are not integrated, you get simplistic media buying, using absurd age/sex demos, which everyone knows isn't right, hence the theoretical appeal of addressable media buying.
Hi Ed, I agree.
This is a very big question and I suspect we will never get to the optimal targeting, partially because it would be a big infringement on privacy. But we can certainly do better than we are doing now. There is more integration. Could there be even more? Sure. I think integration will continue and silos within companies that contribute to the inefficiencies will have to break down. While segmentation is better, it is hard to standardize at this time and there are still unknowns within segmentation behavioral profiles. But the compilation of various data sources and the application of machine learning and AI could make it better. Small steps but some progress.