DVRs could cost networks $600 million in 2007

[morac]

Based on data that states that people with DVRs only timeshift 40% of the time, networks say that they should get partial credit for commercials viewed by DVR users. Advertisers say that DVR users should not count at all even if many do view "live" TV.

Broadcasters stand to lose an estimated $300 million this year and $600 next year under the current system which states that only live (non-DVR) viewers be counted when determining advertising revenue.

Article: http://adage.com/mediaworks/article.php?article_id=113668
Analysis: http://arstechnica.com/news.ars/post/20061211-8394.html

[HDTiVo]

I doubt the networks will settle for this next year.

But there is another way of cutting it...even if only 49% of views are accounted for, to the extent advertisers find value in the DVR 51% (say 27%), networks should be able raise prices to extract much of that 27% anyway. The measurement and pricing is less precise that way, but my point is not all is lost either way.

[True Colors]

How do the broadcasters have any way of knowing if you are watching their show live, or if you are recording them with your Tivo?

TC

[Jabberer]

Ahhhhhhh - that explains why some shows are giving extras (via the web) only to live viewers. SciFi, for example, has Battlestar Galactica stuff posted right after the show for a "limited time" because "you watched it live". I'm thinking The 4400 did something like that last season too. I wondered why.

[Einselen]

Ahhhhhhh - that explains why some shows are giving extras (via the web) only to live viewers. SciFi, for example, has Battlestar Galactica stuff posted right after the show for a "limited time" because "you watched it live". I'm thinking The 4400 did something like that last season too. I wondered why.

1 vs. 100 has chance to win $10,000. Treasure Hunters had their call in/text in/log in competition. CSI: Miami, IIRC, has the CSIQ (get it, IQ, haha).

[warrenn]

Ahhhhhhh - that explains why some shows are giving extras (via the web) only to live viewers. SciFi, for example, has Battlestar Galactica stuff posted right after the show for a "limited time" because "you watched it live". I'm thinking The 4400 did something like that last season too. I wondered why.

I wonder what percentage of people go to the web like that. I've watched many shows live that have some sort of web poll or whatever and I've never even considered going to the website.

[Einselen]

I wonder what percentage of people go to the web like that. I've watched many shows live that have some sort of web poll or whatever and I've never even considered going to the website.

I always enter stuff via the web. Call me cheap but 10 cents (soon to be 15 cents) per message to enter is too much for me. Granted though also the distance between my bed/sofa, tv and computer is only matter of 1 hop, but even at home I still enter online (mostly because I have my laptop somewhere near me).

[Justin Thyme]

I doubt the networks will settle for this next year.
There is nothing particularly new about ad skipping. The traditional way was to go to the fridge for a beer. Jumping to another channel was another.

The Adage article was interesting because it illuminates the crumbling world of old forms of measuring the value of ads based on what time of day they air, and (as a reaction to channel surfing), closer estimations of how many minutes of which shows were actually getting watched. The question they are trying to answer is- how likely is it that my advertisement is actually being watched by my target audience?

What is new is that whereas STBs could measure advertisements that were not avoided, this is much more practical with DVRs since some such as Tivo have two way communication. Since it is possible to accurately measure the success or failure of advertisements to complete an impression, advertizers will more and more demand such direct evidence of performance to see that they are getting bang for their billions of bucks.

That's billions worth of motivation.

I don't see how the networks or the other live media distributors like cableco's have much choice in the matter. It is in Ad time purchasers interests to reject the tranditional statistical inferencing techniques and demand direct measurements of both DVR and non DVR Ad skipping.

Regardless what Tivo's fate is, the DVR genie for ad skipping is out of the bottle. Consumers know they have a choice, they like it a lot, and they will actively seek it in greater and greater numbers.

How will media distributors adapt to an environment where they have far less power to coerce viewers to watch an advertisement?

Co-opting DVRs is being tried- the worse the DVR, the less the ad skipping, the more valuable the advertising time. I just don't see that any of these forms of coersion has any long term hope of survival. You can imbed popup and crawling text banners to guarantee impressions, but you are going to lose audience. This is even more pronounced in an atmosphere of much greater choices of content.

In the end, the distributors are going to push back and say- Hey. Customers don't have to tolerate the coersion anymore so they won't. We have no choice but to make nice with them about advertisements. If the viewer FF's or surfs away from the Ad, it's not our or anyone's fault for providing them the technology that allows them to do that.

It's the advertiser's fault for not creating a message that anyone wants to pay attention to. Advertising techniques have been honed to keep the channel surfer from flipping. I think madison avenue can do better than giggling female breasts during half time, or crushing beer cans on heads. It's not like what is called for is not altogether new- Long ago, we saw cool visuals ads like the Apple 1984 spot. Humor as with the "banned" Tivo advertisement that resonates with viewer anger against network executives... But these were few and far between because it really could not be proved that such ads that did not rely on coersion were any more effective than traditional ads.

What's different is that DVR vendors can now prove when advertisers are doing a good job- which ones are more effective and combatting the advertisement avoidance strategies.

I think the Nielsen mechanisms of infering contact time with target audiences is fundamentally being questioned and that we are going to a pay for performance model based on hard numbers that come from actual measurements of viewers watching advertisements.

[samo]

I think the Nielsen mechanisms of infering contact time with target audiences is fundamentally being questioned and that we are going to a pay for performance model based on hard numbers that come from actual measurements of viewers watching advertisements.
Nielsen mechanism is completely outdated. Modern technology allows much better measurement than statistical sample of average family. For example DirecTV has interactive application that allows you by push of the button to see what other DirecTV users are watching in real time. Application gives you top 12 shows in categories like national, regional, sports, movies, news, kids, etc. It changes show standing in real time and is based on how many DirecTV receivers are tuned to the particular channel. With 12 million subscribers this type of rating is by far superior to "Nielsen families".

[Justin Thyme]

Does DirecTv/ other Murdoch entities Sell Ad time by performance using such DirecTv measurements?

[samo]

Does DirecTv/ other Murdoch entities Sell Ad time by performance using such DirecTv measurements?
I have no idea, but it would make sense if they did. I just use this interactive feature to check for most popular shows.Sometimes it points me to the show I never knew existed, sometimes it makes me wonder about taste of general public. :)

[zalusky]

What seems to be missing from the equation is all the ads that are viewed and ignored.
How about using the DVR to figure out which ads we might be interested and investing in that. Kind of ad technology based upon suggestions. I would be less offended watching an Ad if its something that I might want.

The idea that you have to stand out there be blasted by everything like an ethernet wire is outdated and promotes people to tune it out.

However if you knew that the only stuff that gets through is stuff you might actually buy they would have a huge increase in effectiveness and payback.

Ads could be dynamically pulled from a suggestion pool during commericial and not the same ad for every person.

They need to move on.

[Justin Thyme]

Certainly, much more interesting things can be done. I am interested in a huge number of products but personally have zero patience sitting at a computer waiting for a downloads of video or a flash animation off the internet. If these sorts of things were loaded up onto the Tivo I would look at them if they were interesting. Heck- I'd even buy some of them direct.

I'd pay even more attention to them if I got some discount off my monthly rate, or got Tivo Points for it. And they can verify a human is actually watching by asking questions.

Samo- the main thing is that the DirecTv system know

What the time codes for the commercials are
Which commercial it is
How much of the commercial was watched
When and if the commercial was dumped- either by channel change or FF.
Some idea of the market segment that the viewer is in.

The DirecTv system described is too coarse grained for the kind of performance measure that is now possible (at least on Tivos). Well, the last item maybe a little space age. I am not sure how accurately they can figure out which viewer is watching. Certainly they can model the customer account as an aggregate, and cluster that into subgroups by viewing patterns. I know they have some patents on user identification by behaviors like button press sequences and speeds, but they aren't asking users for explicit marketing information so maybe that sophistication is way off.

[samo]

Samo- the main thing is that the DirecTv system know

What the time codes for the commercials are
Which commercial it is
How much of the commercial was watched
When and if the commercial was dumped- either by channel change or FF.
Some idea of the market segment that the viewer is in.

The DirecTv system described is too coarse grained for the kind of performance measure that is now possible (at least on Tivos). Well, the last item maybe a little space age. I am not sure how accurately they can figure out which viewer is watching. Certainly they can model the customer account as an aggregate, and cluster that into subgroups by viewing patterns. I know they have some patents on user identification by behaviors like button press sequences and speeds, but they aren't asking users for explicit marketing information so maybe that sophistication is way off.
Actually my post wasn't about DirecTV ability to measure commercials effect (although it is very possible that they have this capability). I just wanted to mention that even consumer oriented interactive application is more accurate than outdated "Nielsen family" system.

[Justin Thyme]

It is interesting that advertisers spend billions but are not being informed how well or how poorly they are doing with their very expensive investments. In many respects, it is not in the media distributor's interests to let the advertising customers know they are being fleeced.
Nielsen mechanism is completely outdated.
Actually, they appear to be working on ratings for commercials. But the nielsen moniquer "Commercial ratings" is misleading.

Networks tested the system and one opted out because of serious errors in understanding when commercials were actually playing.

It is more than a little absurd to produce "Commercial ratings" if the granularity of your data is minute by minute. If tens of thousands of users are switching away at the same place in a commercial pitch- that information is gold. Nielsen can't provide that. But it's worse. If they know the user surfed away at a particular minute, was that the beginning of the minute before the commercial played, or at the end of the minute after the commercial played? Assuming that the user patiently waited for the commecial to complete is foolhardy.

The question is, why guess if you actually can know exactly the second when they switched?

Historically, Nielsen's methodology has been indirect- measure the shows that are being watched and you know what commercials are being watched. But with surfing and Fast forward, eyeballs in front of a show are not eyeballs in front of commercials. Users could be surfing away when the commercial comes on, or FF'ing.

Slightly tweaking the measurement to indicate minutes watched of a show was a response to the surfing issue. Attempting to correlate the span of time when a channel was being watched to the commercials that were and weren't played during that time period is all "commercial ratings" is. So twenty years after channel surfing became a national pastime, Nielsen is able to identify some of the commercials being watched but without much reliability due to weaknesses in their data on run time of commercials. Of course, this mechanism is useless for DVR viewers.

And ad buyers and sellers know it.

"I don't think there's anyone out there who thinks that Nielsen has a full grip on this," acknowledges Alan Wurtzel, president of research and media development at NBC... (source (http://publications.mediapost.com/index.cfm?fuseaction=Articles.showArticle&art_aid=46406))

Rogers sent a letter to the players involved, proposing that Tivo's "second by second" data be incorporated into future measurements. What would be especially interesting is if as part of this, ad sellers provided guide data to when commercials play.

The trouble is, if Chevrolet is not demanding to know how much of the ads are being watched. Ad agencies could create a particularly awful ad which induces the Surf or FF reaction, and the Ad sponsor would have no idea. So sponsors are being fleeced as well.

Such performance data will allow advertising techniques to adapt better to the new opportunities that DVRs make possible.

Recently, I needed to buy a family sized vehicle and would have liked to see walk through demos of the various cars fitting my needs. Of course, a 30 second spot shooting the car will silly camera angles would not have done anything for me.

Would Chevrolet buy 5 minutes of time on an obscure channel in the middle of the night if thousands of potential customers DVRs would record it and insert it in place of ads for products the customer is not interested in? You bet. Would sponsors pay for the bandwidth so that DVRs could download them via the net? You bet. When unit sales per month are in the 10s of thousands, you are hitting a substantial percentage of those potential customers with an exceptionally low cost, highly immersive and potentially interactive advertisement.

[jrinck]

Pay ME, the viewer, to watch the ad. I'll be happy to sit through your 30 second commercial if you credit my Tivo/Cable/Sat bill twenty five cents. Credit me a dollar and I'll even let you send me a brochure about your product.

I'll be happy to watch commercials if it means a zero bill each month. Heck, I'll watch even more if I got a nice check at the end of the month, too!

[samo]

Pay ME, the viewer, to watch the ad. I'll be happy to sit through your 30 second commercial if you credit my Tivo/Cable/Sat bill twenty five cents. Credit me a dollar and I'll even let you send me a brochure about your product.

I'll be happy to watch commercials if it means a zero bill each month. Heck, I'll watch even more if I got a nice check at the end of the month, too!
Why would anybody want to pay YOU $30/hour to watch commercials? Some people actually work for much less than that.

[Justin Thyme]

Well, maybe give out Tivo points or a reduced Tivo bill for participation in a trial.

Like ask the customer to indicate products that they may buy in the next 3 months, and see if the customer takes a look at any of the information provided.

[dmdeane]

The DirecTv system described is too coarse grained for the kind of performance measure that is now possible (at least on Tivos). Well, the last item maybe a little space age. I am not sure how accurately they can figure out which viewer is watching. Certainly they can model the customer account as an aggregate, and cluster that into subgroups by viewing patterns. I know they have some patents on user identification by behaviors like button press sequences and speeds, but they aren't asking users for explicit marketing information so maybe that sophistication is way off.I have this vague memory about some kind of TiVo patent, or maybe it was someone else's patent, involving the use of RFID tags and TV watching; anyone remember anything about this? It would be one solution to figuring out just exactly who is watching the TiVo at any given moment; obviously we aren't all going to start wearing TiVo RFID tags; however, I could see TiVo stealing Nielsen's lunch money by creating their own TiVo TV rating service where they get select (ie, voluntary) TiVo subscribers to wear an RFID tag so they can log their TiVo use, much like how Nielsen currently gets TV watchers to manually log their TV watching habits. Currently TiVo users aren't quite representative enough of the general TV audience to replace Nielsen, but they do represent at least a highly valuable demographic segment which advertisers would be very interested in.

[Justin Thyme]

Right- Tivo's patent. My take on it was that the RFID part was mostly to carry authentication to a distant location (EG allow MRV and play on any Tivo unit so long as it is being controlled with a remote that is authorized to play the show). I think user identification was through heuristics like the timings and key sequences we uniquely use with remotes. That's not hard to model, and processing a fast mask to identify a user could be calculated not on the Tivo but on a Tivo server.

(Patent) (http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=1&f=G&l=50&co1=AND&d=PG01&s1=tivo.AS.&OS=AN/tivo&RS=AN/tivo)

Penetrating TCF analysis of what it all means to you and me. (http://archive2.tivocommunity.com/tivo-vb/showthread.php?t=272541)

[BruceShultes]

One of the main reasons for fast forwarding through commercials is one company shows the identical commercial several times during the same show.

This has happened at least since VCR's became prevalent, so the effect on commercial viewing has not changed recently.

[vman41]

How do the broadcasters have any way of knowing if you are watching their show live, or if you are recording them with your Tivo?

TC

Nielsen has boxes that pick up digital watermarks in the audio stream. The gizmo that inserts the watermark has been a source of trouble with my NBC affiliate.

[ZeoTiVo]

I had recorded some toy fair show from HGTV to see my kids reaction to the different products and get some ideas of what they would like toy wise.

This ad thing could go to a whole new level if "a toy show" was recorded onto the TiVo around Thanksgiving and viewers could hit thumbs up or down at various places. The thumbs up could (after privacy policy and agreement to such)result in an email to the account address or snail mail to the house address with the reults and local store coupons for the same.

selling merchnadise and generating foot traffic into a local franchise with a coupon to log the visiits. Pretty easy to gauge the ROI on that and extremely useful to Mom and Dad as well as to the stores who with a better gauge on interest by zip code could shift inventory and so forth.

[Stormspace]

Well, maybe give out Tivo points or a reduced Tivo bill for participation in a trial.

Like ask the customer to indicate products that they may buy in the next 3 months, and see if the customer takes a look at any of the information provided.

I've participated several times in Neilsen research and each time I considered it my duty to keep my shows alive. That was my payment.