Commentary
Programmatic TV Ad Buying Will Never Work
- by Ari Rosenberg , Op-Ed Contributor, September 8, 2016
Pre-new media, the definition of a “publisher working” was primarily defined by how many people consumed the content a publisher produced in any media, and the depth of this consumption.
Under this definition, publishers knew that if they invested in doing their job well, they would earn more ad dollars. Ad content would also be held to this same working definition: Did the ad creative earn consumer attention, and to what depth?
The beauty of this old-media definition of working was the proper alignment of incentives. Whether you were a publisher or an advertiser, investing in better content produced deeper connections with consumers. As a result, publishers made more money, and advertisers felt better about spending it.
When new-media publishers arrived, driven by the frenzy of venture capital dollars, these publishers agreed to deliver unrealistic ad revenue goals despite having no established record of “working.”
advertisement
advertisement
When these new-media publishers started making sales calls, they learned quickly that while they had a large number of total audience members, the depth of their consumer connection did not compare well to old media. The old-media definition of a publisher working placed new-media publishers at the end of the ad dollars line.
This was a problem, because invites to their IPO parties had already gone out, so new-media publishers had to figure out a way to cut the line.
They did so by agreeing to a new definition of “working” that new-media buyers rewrote for them. Publishers “working” was no longer defined by the depth of the consumer attention their content earned -- but rather, how quickly a publisher could transfer this consumer attention to an advertisers’ Web site. New-media publishers were desperate to grow revenue quicker than they could earn it the old way, so they agreed to this new definition of media value, and ad dollars started to flow their way.
This was a brilliant move by new-media buyers, who now could easily drive down prices by comparing how their ad creative worked from one publisher to another, while ridding themselves of the burden of investing further in creating ads that worked. As long as prices dropped, their ad creative would “work better.”
This new "working" definition, however, created a misalignment of incentives.
Publishers were now incented to produce content “just good enough” to borrow consumers’ attention long enough to flip it instead of owning it. Ehow.com was a perfect example of this new paradigm. Ehow.com built its whole business on having consumers visit its site just long enough to transfer them elsewhere via search ads run on the site. So Ehow.com could collect its Google Ad Sense toll, and users had to travel further to get what they were looking for.
The key misstep for new-media publishers was not using the ability to track consumer actions with their own content to prove they were “working.” This would have kept incentives aligned so that consumers were treated to ever-improving content from both publishers and advertisers.
Instead, we have what we have today.
Programmatic is billed as media’s future, but don’t be fooled. It’s not about automation. It’s not about giving TV networks and cable stations more insight on their audience to help them sell more advertising. It’s not about attaching data to increase the value of ads sold during low sell-through time slots.
Programmatic is all about ad buys based purely on audience targeting. By default, content value is eliminated, and the new-media definition of a publisher “working” gets further entrenched.
TV executives, however, are not desperate. They stand today where they have always stood: at the front of the ad dollars line, with ad scarcity and highly produced content in their pockets. They casually look back over their shoulders at the other media behind them. Their smiles say, “Settle in, folks, we are going to be here a while.”
Programmatic ad buys, by definition of how they are working, will always drive prices down and decrease overall spending. TV executives will pay programmatic lip service to sound like they are evolving technologically, but they are not going to open the door to their inventory wide enough for programmatic to work. Nor should they.
Food for thought, Ari. Well said.
Agree with your definition of "work," Ari, and, yes, the dollar-value of well-produced content has taken a beating as programmatic grabbed budget, but keep in mind the definition of 'It's a dog's life,' when thinking about digital media.
We say one year for a human is the equivalent of 7 years for a dog, and so it is with the Internet.
You can stop to consider Television's position as being at the front of the line, but the speed that time-spent with newspapers, magazines and radio went the way of parchment, has me thinking that TV executives are positioned to be the last media lemmings to go over the cliff.
I am not the first one to say this, of course, Ted McConnell most recently published something along these lines a week ago, but I can not agree with you that this will "never work."
That is the 'misalignment' in your argument.
Evolving technologically is going to change the definitions now associated with programmatic.
The digital traces content/ people/ devices weave into data online can inform broadcast TV.
The opposite is not true.
And that's what still needs to be figured out.
As it should be.
Lubin, radio listening never dropped off a cliff, it merely shifted gears when TV usurped its in-home program entertainment functions and shifted to a service and background listening mode, with most of its consumption now done in cars. As for TV it's not just a function of audience---and it will be quite some time before TV loses its huge edge over digital video where reach and time spent are concerned---a very long time, if it ever happens. But there's more to it than that. Advertisers ---and audiences----crave the kinds of content TV offers as it generates involvement and image enhancement benefits that go far beyond crude audience counts.Digital is just getting its feet wet in this vital area.
Why is programmatic TV moving so slowly? The answer is obvious. Programmatic buying, as executed in digital, simply doesn't fit the linear TV advertising and ad selling model. It's not just a question of compelling TV to accept digital's direct response- driven metrics, as if that's an inevitable outcome. Rather, digital and TV must adapt so each uses the most relevant approaches for their mutual good. That means that programmatic platform suppliers must learn what TV advertisers need and what TV ad sellers need, rather than trying to force them to do things the digital way. If not, then TV will get along just fine without programmatic.
Ed, there is another factor with broadcast and cable TV that advertisers crave.
Given that they are linear, an advertiser can buy an ad at 9pm the night before their brand new product or store opening or sale, and know that millions of people will have seen it simultaneously. That is TV (and radio) can have a virtually immediate impact on demand. Online audiences, being on-demand, accumulate over time, which makes the staging of launches, promotions and sales problematic rather than programmatic. While increasing numbers of people are preferring 'on-demand' as their viewing option, this doesn't always help the marketer.
Thanks Lubin for challenging my beliefs -- I see your well made points and John and Ed, thanks for supporting my beliefs and adding wisely to them -- John I had not thought about that point of timing before a product launch -- that's a great point. This is a very complex topic but I like to look at things through a very simple lense and the simple truth is no publisher wants to sell their inventory programmatically -- it's a classic chicken vs the egg scenario -- they do it because the ad budgets are moving there and the ad budgets are moving there because publishers are doing it -- TV has always had the leverage to not do what buyers want them to do and I suspect as my column indicates, they're not gonna budge -- buyers fear missing out far more then us sellers would believe and they will balk and spend their ad dollars the way the networks dictate not the other way around.
Ari, I must respectfully disagree with a lot of this, although the point of view is not uncommon.
Programmatic did not drive prices down. Publishers creating inventory from thin air did that. In fact, publishers support audience targeting by retargeting their own audiences, right?
Audience targeting does not demean content. It just gets advertisers what they wanted in the first place. Content had been a surrogate for audience since the beginning. Look at the flow of a media plan from advertiser to media. It starts with audience, and great content gets more audience but not necessarily aligned to every advertiser's needs.
Does context help. Yes. Tons. That's why Brands want both.
Auctions only drive prices down when the buyers can't inspect the goods. If they can, the prices go up if the goods are good.
Clicks were indeed the definition of "works" for direct sellers, but they never were for Brands. I have Brand after Brand tell me they don't care about clicks. In the early days, clicks were so irrevalent that the agency would not even code a click tag in the ad! And true, clicks are great for DR, but remember that the in the total world of retail, only 6-9% is online. The other trillion is driven by the top of the funnel. There was never any idea that clicks drive commerce. There were just publishers who wanted to serve the campaign objectives of their DR focused customers, which is fine.
Having created an exchange for linear TV, I can tell you that programmatic TV advertising can work for seller and buyer for the simple reason that it creates a transparant distribution channel for a high quality medium. I would predict that programmatic would work better for TV than online precisely because oversupply is not a problem. Precision matching of audience with need is a killer strategy.
PTV is in fact a lot about workflow and (you did not mention) access to the medium. Broader access to the medium creates fill rate, which drives prices up because media value goes up when niche audiences can meet niche products. How many customers does Google have?
So, yes. PTV will take some time, but only because of the natural vaguries of Broadcast communications. But, soon, data will prevail. We will be able to evaluate spots, and pay a fair price for them. Remember that buying groups are complicit with all this. Maybe they should buy the best spot, not the cheapest. Its up to us to be able to show what's good. Shame on us if we can't.
Ted thanks for weighing in -- your experience brings so much to the table and I read your comment multiple times and with an open mind.
I am not suggesting reaching the right audience isn't the underlying goal of an advertiser but remember that whole experiment of that classicly trained violinist who was ignored while playing at a train station -- and then played in front of thousands in a theater the next night where people paid over $100 a ticket (or something close to that) -- programmatic puts the spotlight on data and hence audiences and heaviliy discounts the impact the theater setting has on how the ad message is perceived (I am sure there were plenty of people in that train station who from an audience perspective, matched those in the seats at the theater).
Brand clients talk about how they don't care about clicks and yet they buy Google as you mention -- on a cost per click basis. They don't want to use clicks as a metric to define performance I get that but they will continue to define performance on their terms and my underlying argument in this column was that buyers/clients shifted the definition of "working" to drive prices down and "programmatic ad buying" helps further facilitate that definition putting publishers in the small chair at the leverage table -- and while many many web publishers had to accept those terms TV networks don't.
Thanks again for your inisght -- I read your column religeously so I was flattered you read mine.
I feel compelled to point out once again, that there is quite a bit more to branding advertising than merely maximizing the number of targeted eyeballs "reached" at the lowest cost. Aside from reach, which programmatic TV can't handle for "linear TV" , there are perfectly valid concerns about media environment, context, compatability and merchandiseability that programmatic TV advocates simlpy ignore or, since they have no way to handle them, treat as minor details to be worked out--somehow--at a later date. That isn't going to sell. Sorry.
As for targeting TV buys in a better fashion, if the "data" people bothered to look at what data is available for linear TV, they would find that old fashioned demographics---say age within income---predict about 75% and in some cases, more, of the product use metrics. Moreover, there isn't the great variation from one show to another that is presumed where most program genres are concerned. If you substituted product use data for demos as a way of profiling TV show audiences, and the sellers allowed you to cherry pick their schedules---fat chance----except for very youthful or super affluent or ethnic targeted brands , you won't get the huge improvement in targeting efficiency that is promised so long as your daypart and network type mix remains constant. Worse, if you do, somehow, improve your targeting efficiency--say by 10%---during your first buy, this will probably be a one time event as the sellers will up their CPMs on any shows or genres that are in high demand.
Old Media Pros like Ed can spin all they want but the facts vare the FACTS.
If as you all state Old Media works then consider:
1) No where in all of your statements and analysis is the thought of the user and viewer and what the valur prop would be for them, therefore a fundamental flaw.
2) How does this apply to HBO? or Nerflix? Hint: think value prop for the user and viewer.
3) Why are the demograhics hitting the geriatric stage? Where will growth come from? Selling less audience for more $$$$ is not a sustainable model for BRANDS.
4) Lack of transparency is the reason why Programmatic is handcuffed in TV. Holding up the old business model as long as the spin continues like musical chairs is no way to plan for the FUTURE.
5) The mere statement that Programmatic will never work is like saying technology will never work. Good luck with that.
So what we have here, to borrow from an old movie is, "a failure to communicate". On the one hand we have the avid digital evangelists, who operate on the assumption that everything about "linear TV" and the way it is used by advertisers is totally wrong, hence unworthy of consideration, let alone accommodation. Arrayed against these brilliant digital futurists we have the pro-TV "spinners" who are actually not very bright and set in their outmoded ways. Obviously the latter will soon be relegated to the elephants' graveyard as TV becomes totally digitaized and embraces the digital mantra. What's more, any factual evidence that the "spinners" cite is crap and should be ignored.
Sort of reminds me of our innane televised political "debates" where nobody hears what anyone else is really saying and there is no give and take----just lots of slogan chanting and not so veiled personal barbs.
Ed, I hope you didn't take offense -- I think Ted offered some great thoughts to consider even though I still stand by my thoughts that oppose his -- as for Leonard's comments -- they are mean spirited and he is probably having a bad day.
Keep your thoughts flowing here Ed (and other posts) -- your contributions are greatly appreciated certainly by yours truly.
Thanks, Ari. I had no problem at all with Ted's comments, though he and I probably disagree regarding the probable outcome and the relative importance of "data". As for our resident, slogan spouting, digital evangelist, it's a shame that he can't engage in a true debate, where ideas are bantered back and forth and the "other side's" viewpoint and evidence are respected rather than being dismissed out of hand. The guy seems to be quite intelligent, judging by his posts on other subjects, but this is, frankly, not evident when he adopts the role of the young gunfighter looking to make a name for himself by challenging the "old hand". As for the value and merits of such encounters, I leave it to our audience to decide. I have long since abandoned our "debates".
Here's what I see happening in programmatic TV right now: Some stations are pulling direct response inventory and attempting to raise their net income by selling it at higher prices on programmatic. In other words, some stations are attempting to leverage their ad prices back up.
But the only other thing that's sold under the title "programmatic" are local inserts on cable networks - in fact primarily via satellite.
So, none of that is really programmatic - it's just trying to sell discounted TV time to major advertisers. Kind of a mess. And it's really not a very good deal because a good buyer can out negotiate a computer program.
That said, the weakness I see is this: There's no data. Even the supposedly "excellent" data on the web is really poor for programmatic buying. Otherwise, why am I pummelled with so many ads that are so badly (and obviously) targeted via programmatic?
Yet there's a lot of click and website data to use to develop profiles of individuals to sell to advertisers online. With TV theres...um...well... nearly bupkiss.
Hope Ed will correct me if I'm wrong - but there's so little data that can be connected to a TV viewer that calling this "programmatic" is really only taking a hypey term from the digerati and trying to make a ton of money by just calling whatever you create "programmatic".
At bottom, I'll agree with the title. Programmatic TV buying will never work.
Very good post Ari! I agree that audience targeting is the first of the "programmatic" flavors that will impact TV, and my experience has certainly proved you correct in that TV media owners don't need to be in a hurry to shift into this world too fast - the current models isn't so broken that anyone needs to implement quick fixes. I do think that over time we will see some forms of automation and digital/TV cross-buy packaging coming in, but all will take time. Of course, in a $75 Billion business growing at 6% per year, that could amount to a non-trivial amount of spend even if it is small as a percentage of the total.