CW: What is your definition of television?
EB: Television is produced content that is owned by the media company and monetized by advertising. Video is shorter in format and may be user-generated, but is also ad-supported. In contrast, a film, a movie, or a series can be accessed on the same devices as TV and video, but these are monetized by subscription. Finally, cinema is viewed in a dedicated venue and monetized by admission fees. It boils down to monetization models, which is why some of these definitions may converge in the near future.
CW: Where is ad tech headed?
EB: Migration to programmatic will continue. Today, marketers are embracing the basic benefits of efficiency and scale for basic media channels like online display, video. They do this through DSPs and programmatic direct deals. Moving forward, brands will continue to adopt programmatic and will progress to more optimized one-to-one approaches. We are definitely headed toward programmatically traded media in other channels like TV, print, outdoor, etc.
advertisement
advertisement
CW:If you were at a television network, would you start the sales shift to programmatic? If so, how could you maintain the CPMs?
EB: I would certainly start the sales shift. There are many excellent reasons to shift to programmatic, but the most immediate would be to [help] figure out the space, possibly even to contribute to shaping it while it is still malleable.
Any television network that ventures into programmatic for the first time a year or two from now will be riding the coattails of their competitors. The CPMs will do exactly what they did in online display when that business became programmatic: they will dip a little short-term, then they will recover.
Of course, every programmatic platform supports floors, and floors can be used to prop up pricing, but I don't think CPMs are the issue. If CPMs drop a bit while volumes climb, net revenue won't be impacted. Programmatic is sure to bring a whole set of new advertisers to TV, while it brings engaging TV ads to 1:1 advertising. I think everybody wins from that.
CW: Give me some predictions for how the media landscape will look five years from now.
EB: In the next five years, we will continue to see a shift in ad tech as it relates to marketing technology. Companies are expanding their capabilities to become integrated marketing platforms. Enterprise software companies like Oracle, SAP and IBM are increasingly becoming involved with helping CMOs in a way similar to that of ad tech agencies. Oracle’s acquisition of Datalogix back in January is proof positive of this trend toward all-encompassing marketing solutions.
Similarly, ad tech solutions platforms like ours are strengthening our marketing arm on both sides of the equation. While we partner with many advertising
and media agencies, we also work directly with brand marketing leadership, serving them as a programmatic marketing platform.
If TV CPMs "dip" with the theoretical advent of 100% "programmatic" buying---and I assume he means real programmatic buying with 95% plus of all GRP inventory at play not "computer assisted" human buying tests----and this is offset by more "volume" for the sellers, doesn't that translate to more ad clutter------which means less impact per ad? Or aren't factors like ad clutter and impact part of the "programmatic" process? Just asking.
Good question. I will forward to Eric for comment.
I'm also intrigued to hear the claim that programmatic will bring a whole set of new advetisers to TV---I assume this means direct response advertisers, but maybe he means companies as yet waiting to be formed. Also, the reference to "more engaging" ads is interesting as this probably refers to "addressableTV". So, thanks to programmatic, advertisers will be able to single out only those viewers who are really interested in what they have to say, therefore the commercials will be more"engaging".
It's all so pat----until you ask yourself, what percent of the total GRP pie consists of viewers who are in the market for an average product---assuming that it is really possible to determine this and find a database to guide the programmatic computers to said viewers? If the answer is 50%---and I may be being generous----what happens to those non-prospect GRPs? The counter to this is the assumption that it will all even out---meaning that the prospects for one ad campaign may not be the same people as the prospects for another advertiser. Yet most marketing studies tell us that roughly 20% of the population spends about 40% of the amount taken in by advertised products and services, while another quintile, accounts for only 5%. So, maybe things wont even out after all.
Hi Ed, Eric has responded and asked me to post for him:
Hi Ed, thank you for reading and we appreciate your thoughtful questions. I’d like to address some of the points that you raised. To your question about volume: right now, we see a certain number of ads, and some percentage of them consists of house ads, the ones that advertise another show on the same network that will come on later. As volume increases, these unpaid slots will be consumed. This eventually leads to increased competition for existing slots, and two things happen. Prices rise – that is, recover from the dip – and improved economics lead to generation and broadcast of more content (with the same number of commercial slots in it.) Of course, more content means more fragmentation of the viewing audience so advertisers will rely more on programmatic buying platforms to find their audiences. It’s inevitable.
Ed continues -
Regarding a whole new set of advertisers, yes – we might expect that direct response advertisers could consume more TV air-time using interactive TV. However, I expect that programmatic TV will evolve similar to the way that programmatic video already has – that is, mostly for branding. Branding and direct response are two different marketing tactics and most advertisers use both. What I expect will bring new advertisers to TV will be the 1:1 nature of RTB programmatic. I agree with you that TV advertising will be increasingly attractive to brands with more targeted audiences.
On the point of too many GRPs being exposed to ads they’re not interested: 1) GRP will become irrelevant, and 2) 50% is way too high. GRP is a metric that will not survive the transition to programmatic. It’s only useful in measuring the portion of broadly defined audiences that are reached by an ad through limited channels. In the world of programmatic, viewers are selected one at a time. Campaigns are very specifically targeted by audience attributes as well as by content and situation. With these selective impressions, GRP becomes meaningless. If we use response rate as a proxy for the number of people who are interested in an ad, let’s call it the relevancy rate, then the prevailing relevancy rate is in the low single digits. The real interesting point is that, compared to other advertising techniques, programmatic improves the relevancy rate quite a bit, often doubling or more. To me, that is a worthwhile improvement.
Thanks, Ed, for the conversation! It was great to give into more depth on what we at ChoiceStream feel is a fun topic.
Thanks for your replies, Eric and Charlene. I must note that I find some of the assumptions about the application of programmatic to TV---in the same manner as it is used for digital----to be problematical. For example, what database is going to allow programmatic trading desks to single out individual viewers and send them tailored ads? No such thing exists or is even in the offing.
There has been a lot of talk about melding "big data" TV set usage ratings with third party product purchase and related data, then mashing the resulting indices onto the Nielsen ratings as if this singles out a marketer's prime prospects---but it doesn't. All you get is a tonnage buy, modified slightly by the weight of the indices and time charges.
As regards targeting, a typical branding advertiser wants to reach not only his current customer base but also those who buy other brands. Also, it matters whether these are heavy, moderate or light users of the product. As a result, most brands assign weights to various sectors of their category's user base and strive for varying levels of reach and frequency against each group. How is this aspect to be dealt with when there is no database that combines all of these elements for the TV trading desk to evaluate?
Finally, and this is the biggest problem, the ongoing assumption seems to be that the TV networks and cable channels will give up control of their GRP inventories and allow advertisers to cherrypick within their schedules at the same time driving for the lowest cost per targeted viewer. This is simply not in the cards, for the major sellers, leaving the programmatic option to be applied mainly to the smaller fry and their "non-premium" GRP inventories.
I happen to think that computer based systems can be a big help in planning TV buys and in allocating the time made in upfront corporate buys to individual brands. Whether programmatic will ever dominate the actual TV buying process will fit in and evolve in the same manner as seems indicated in digital remains a big question mark.
Correction on the last sentance. Delete ---"will fit in"---
Ed - Love your comments. There are very practical problems with all the programmatic theory. But it seems the market has been giving the proponents a free ride based on the theory "we're applying a successful digital approach to TV"...
But there are sooooo many problems with that theory. Probably the first being: How do you know it works in digital? CPMs are horrible because impact of ads is so bad.
And there's no free ride in media. There's entirely reasonable theory that TV rates are (in general) set at what the market will bear. No clever re-selling mechanism that "takes over the industry" is going to magically reduce their prices.
I made that observation about GoogleTV when their salespeople were pitching so hard about what they could deliver. And GoogleTV was a bust - a bad bust in most cases.
Similarly, programmatic will be a bust (I believe). But there probably won't be a moment when we can realize that. Because so many powerful political and financial players in the media are making a big commitment to it that they won't be able to admit error - they'll find a way to "declare victory and go home".