In about 2003 the editors at Faultline got their first lesson in Addressable Advertising or Targeted TV advertising and we were blown away by it. The premise is simply, if you have 15 million people watching a TV channel, and only 5 million have a driving license, then if you played a car advert to the other 10 million, it would be a waste of money. Just as many people would respond to the advert if you only showed 5 million ‘impressions,’ to borrow a word from the world of internet advertising, than if you played it in all 15 million. So why not ONLY play it in those 5 million homes and sell more advertising to play in the other 10 million homes, increasing your revenue.
We were led to believe that advertising could be sold at slightly cheaper prices, but sold ten times over, so that something like 8 times the amount of advertising revenues could be raised. But it’s not as simple as that, and more or less nothing definitive has happened in this area.
This week in the run up to the IPTV World Forum, recommendation engine player TV Genius has published a white paper on the subject entitled ‘Enhancing Targeted advertising with Content Discovery’ and it’s fairly enlightened, and states many of the problems associated with what we shall continue to call Addressable Advertising, although throughout the paper TV Genius prefers to call Targeted advertising.
Flicking through its pages, there are a lot of good points, but at Faultline, over the past 8 years we have heard lots of good points too in favor of and against changing the world of advertising, and we thought we’d use the trigger of this paper to talk some common sense, and to explore some possible futures in TV advertising. We will unashamedly borrow some of these points from TV Genius, hence the credit, but we’re not sure we reach the identical conclusions (or any conclusions).
The first thing to ask is why hasn’t advertising become this fragmented, hyper responsive and aware environment which might bring in ten times the number of campaigns? And that answer is simple.
Most advertising is sold by what we shall call Networks, companies which create TV channels, either designed for free to air environments, which we shall call Broadcast Networks or for pay TV platforms, which we shall call Cable Networks.
Most broadcast networks use terrestrial broadcasting or satellite broadcasting and they reach every home in more or less the same condition, with the same adverts. The transmission operator, often a low margin business which serves all broadcasters in a given territory (in Europe), could only offer such a service to all of them collectively, and even then it could not access complex demographics in order to deliver adverts to different homes, the best it can do is to offer regional differences in advertising by splicing different advertising into local programming. This happens and it makes for improved responsiveness, but in a world where internet advertising is taking more ground by the day, it’s not enough. In the US transmission is supplied by a single market broadcaster, and they have even less money to spend on innovation, even though they have carriage agreements which allow some of their own advertising to run.
But in the Pay TV market, the operator who attends to the delivery of the service has margin aplenty and constant capex to consider and often triple play potential. The network ‘ whether it relies on satellite delivery or ancient analog cable or digital cable or digital switched video or on IP carriage, can have either a return path so that data can travel to and from the TV set, or it has a mass storage device such as a DVR, which can store advertising and complex programming or alternative advertising, or it has both.
In a broadcasters market it would be easy to send advertising to a DVR and splice it into programs at key marker points, except that DVRs are delivered as a retail proposition, bought by the home owner, and they are built to standards ‘ so addressable advertising rules would have to be standard from day one, and that’s no longer possible.
Only a Pay TV operator which has the margin to supply a DVR, which is built with specified software and ad delivery capabilities, could pay for and standardize on a method for exploiting addressable advertising.
There are issues, as the TV Genius paper points out, some technical, such as how you deal with the security issues of moving demographic data and program preference data around a network without it upsetting consumers, or being ‘stolen,’ so it can be opened to public scrutiny.
But things are changing and here’s just a couple of the things which are changing ‘ first we are getting more and more internet connected TVs. This will result in more and more Over The Top programming, and this will come over the open internet and will naturally give way to internet advertising models. This is more about interactivity ‘ and there’s a big gap between advertising which targets people that are already interested in the subject of an advert, and the immediacy that a brand gets of people responding directly to the adverts. Both are desirable but one gives operators more revenue, while delivering the same results, the other gives the advertiser clearer ways of measuring campaign effectiveness, and makes the gap between paying for advertising and getting revenues back from it, far shorter. Of course having both is what we should be after, more effective adverts which are interactive. The other way in which things are changing is that IPTV pay TV operators are now reaching the stage where the larger suppliers have 4 million homes reached, some have 3 million, several have 2 million, and quite a few have 1 million, around the world. These are just entering the phased of their development which would make it worth investing in new advertising technology.
Addressable advertising can be addressed in many ways. Either you can place local adverts in nearby viewing locations, so lots of local businesses can buy the local spots, broadening the appeal of TV advertising. But does this get you more money or simply a broader market with more indirect benefits? It can be addressed by content ‘ Motorbikes advertised on Motorbike shows, etc’ TV Genius has a key addition to this which is Motorbikes advertised to people who at some point have watched a Motorbike show, or who are being shown recommendations to watch a Motorbike show because their demographic says they might like to buy one.
This is the entire TV Genius point of its paper, that if you have a proper TV recommendation engine, you can put addressable advertising in a better light.
Because the last dimension, that of looking at what you as a family purchase, what kind of income bracket you are in, how much you paid for your home ‘ that’s the sinister bit that worries everyone. There is a further worry, and that’s called Psychographics, where you advertise to people based on what they believe, or prefer emotionally ‘ pandering to their prejudices and attitudes.
Imagine what the Facebook generation would make of having a presenter wear a red tie in a Labour Household (UK political party) and a blue tie in a Conservative home. It might actually make a difference if not noticed, but people move around and watch TV in other people’s homes and given that the way you vote is not yet available as a demographic (that would be illegal) then some homes vote Republican in Democrat (US parties) neighborhoods and vice versa.
Adverts which rely on your prejudices are an opportunity to get things totally wrong. But it would also lead to a society where people are never exposed to opinions which are different from the ones they already hold. That’s not going to work in a world which embraces diversity in social networks, or is it. Don’t social networks just allow you to hang out with people who think the same way as you and avoid those who think differently? Hmm, another time.
So while addressability of advertising requires investment, and that can only fall on the link in the TV chain that both has money and some basic advertising potential, given that their revenues are ‘not broken’ and that these are the same organizations that weathered the storm of the last recession particularly well, why would they mess with their advertising.
Well as we said before, if OTT channels come along, they can be sent with or without existing advertising, and as the IP protocol is more widespread for TV, both in QoS controlled networks such as IPTV and over the open internet, there will be moves to break the hegemony of satellite and cable, and one way to do that is to attack this space. If traditional pay TV operators are not ready when this happens, things could move against them pretty fast. So the model is not broken, but will break pretty soon, so get ready.
Another issue is that consumers want fewer ads, not more. Initially OTT services will offer fewer adverts of shorter duration, in order to establish the medium. We sense that the last thing that any broadcaster could or would do is show less adverts, so this will become a key ingredient in the shift to OTT services. But if we can’t do that on free to air terrestrial broadcasting, we perhaps can do that on a pay TV network with a return path or a controlled mass storage device like a DVR.
So why shouldn’t the world simply move to a stage where DVR’s are turned into a revenue stream, by broadcasting or uploading advertising to the DVR, in the same way that push VoD is performed, and then splicing these ads into programming, as and when it is needed based on the audience preferences (whch might even be explicit preferences). At first it could be with push VoD programming, currently shown without advertising for the most part, or it could be shown outside of the normal viewing zones, on the EPG for instance.
One of the neat things about ownership of a DVR is that it often self-selects higher income families who can afford high levels of pay TV service. Or does it? It used to, but more and more, as households use TV more and more, lower income families can afford alternatives like the internet and other genuine hobbies less, but insist on a DVR. So perhaps not.
When we shift to using demographics of different households to select advertising we are faced with some real issues. If your home gets adverts for brands you already like, or for brands you are deemed able to afford, surely when you watch next door, and see different ads, you are going to get a Big Brother feeling, you are going to feel defined and classified and fell pretty unhappy about it. You might even take it out on the brands that are being pushed at you, by never buying them. It may turn out that advertising is as much a shared experience as watching the TV and people want it to be more or less the same.
So we were drawn by some of the those ideas of TV Genius, for instance to play adverts in the EPG, showing the advert instead of an EPG entry which shows a program the viewer has already decided to record or has put on Series link (record the series). That way it doesn’t cover anything you might choose to watch, since you have already chosen to watch it. But it went further, suggesting that programs thrown up by its recommendations system might also be used to drive onscreen promotions of other programming. We’re both hot and cold about that idea and it’s the same problem we have all the way through. Should you promote programs that you already ‘know’ people might like to watch, or does that defeat the entire idea of promotions.
Should a Pay TV operator want to push the programs that it has paid the most for, so that it can drive more advertising there, or should it push more programs that are in prime time, or should it promote programming that extends the viewing hours of the person watching (he always turns off his TV at 11.00 pm so advertise the next program). In fact eventually you get to absurd questions like should you re-schedule programs that most people want to watch, to times when most people are watching?
Surely that’s the job of the program scheduler, but you might do that automatically with such a system. But of course all of this goes out of the window with an OTT catch up TV service, which carries new advertising when you watch it outside of its linear playing time.
Will advertising on TV get more and more like the internet, in which case if your ad delivery system knows it owes 1 million ‘impressions’ to an advertiser to fulfill his schedule by midnight tonight, does that advert override all the other adverts on the system? Could we find that installing new technology in the DVR, or on the servers of OTT services, to offer Addressable Advertising, leads us inevitably down that path?
TV Genius points out that the global TV advertising market is worth around $160 billion a year and accounts for more than 40% of the global worldwide advertising spend. It also points to the fact that pay TV operators take less than 5% of these advertising revenues, with the remainder going to the networks. So since TV Genius is a supplier to pay TV operators, it is on the side of some of this money changing hands.
It points out that UK satellite DTH service BSkyB receives around $7.5 billion a year from pay TV subscriptions, but only £300 million from advertising, giving gives it 6% of the UK TV advertising market, even though it has well over 50% of the pay TV market.
What would happen if suddenly Sky could get as much as 25% of the UK’s TV advertising market? Well of course all the best channels that it carried would either a) go bust, b) charge it a far higher carriage fee or c) buy less high quality content, and attract fewer audiences.
So that only works if Sky ALSO simultaneously increases the total size of the advertising market, so that the Networks don’t get short changed, and actually get an increase in the ad dollars they get, but Sky gets some percentage for making this possible. Is that enough for it to take on ALL of the risk?
Throughout Addressable Advertising issues there are always ‘damned if you do, damned if you don’t’ situations emerging, which is why companies who feel that their marketplace is ‘not broken’ are unlikely to spend aggressively here, and instead it will be left to companies which are new to the TV market, not those already doing well. Addressable advertising could become a tool for eroding market share, not shoring it up.
TV Genius concludes in its white paper that a lot can be done with recommendations, as well as promotions which rely on recommendation data, and tout openness to the consumer, so they don’t get sidetracked by security concerns.
We respect that it is trying to promote a style of advertising that might come to be the salvation of advertising at some far off point in the future, when localized advertising mechanisms such as local newspapers and billboards may potentially be dead, and replaced by electronic means.
But we think it is missing the whole picture. The personal nature of TV experience needs to be centered on a personal device. The natural home for personal favorites and program recommendations is on a handheld device ‘ a tablet or smart phone ‘ because each person in a family has to have his or her own record of what they watched and his or own record of making remote control key presses.
As handsets are turned into universal controllers for WiFi aware connected TV sets, this entire paradigm shifts once again, and another screen is available for showing advertising, the handset-TV remote App. That is currently the property of another kind of operator, either the cellular operator or the independent CE manufacturer, and they will have to become part of this equation as preferences follow individuals around and slip back and forth into the cloud and onto multiple devices types, making security a far, far greater issue in the future than it is today.
And as far as we know, there is virtually no dialogue between the modern set top or DVR and the modern handset, and until the mechanisms for that are in place (see NDS article in this issue) the potential for full addressable advertising will never be in place. Just think, what if, not only does your TV show you a different advert from the neighbors, but your handset, when you bring up the EPG that controls the TV set, shows you are a different advert from your father, mother, wife’s handset based EPG. Only then will we have the full picture in addressable advertising, right down to the individual.