How To Bring Internet Advertising to TV—The Long View
by Erick Schonfeld on March 26, 2008

blinkx-ad-hoc-2.png

Over the past couple days, conversations I’ve had with two different video-startup CEOs—Suranga Chandratillake of blinkx and Nick Grouf of Spot Runner—has got me thinking about what needs to be done to make TV advertising as relevant as video advertising. We have a long way to go, but it boils down to two things: 1) replacing 30-second commercials on TV with relevant ad overlays that pop up at exactly the right moment during a show, and 2) automating the buying, creation and placement of TV ads to make it more like buying search ads.

Yesterday, blinkx CEO Suranga Chandratillake dropped by my office and we got to talking about video ads. Blinkx is a video search engine, but it is also building a video ad network called AdHoc that attempts to place contextually-relevant, clickable text ads in a bar above the Web video being watched. For example, in the screen shot of the soccer video above, you will see a text ad for shoes. YouTube is doing something similar with its new AdSense for Video ads. The ads themselves don’t have to be text. They can be banners or logos that pop up, or even new videos-within-a-video under the control of the viewer. The point is that they exist within the main video itself, not after or before it. And they appear briefly at relevant points during a show.

Determining whether an ad is relevant for a video is done with some of the same techniques used on Web pages. Both YouTube and blinkx look at the tags and text surrounding a video, but blinkx actually goes beyond what YouTube does. It uses powerful speech-to-text translation technology to create a transcript on the fly and then matches relevant ads to the words. The ads appear as those words are being spoken. Suranga showed me the transcript-creating capabilities of blinkx, which are not visible to users on his regular site, and it was impressive. He clicked on a word in the transcript and that point in the video started to play. Once you can do that, inserting relevant ads is trivial. He says blinkx can also match ads to related concepts extracted from the transcripts.

What would it take to run ads like this on regular TV? Even if the ads are not clickable, simple banners or graphical buttons that appear in sync with what you are watching would grab your attention. Imagine a Nike logo popping up when you are watching basketball, or cruise ship when someone in a movie mentions the Bahamas. It could be annoying, but not if used judiciously. And it would certainly solve the problem of people fast-forwarding through ads with DVRs.

The big issue would be separating the ads from the underlying video so that new ads could be placed when the show goes to cable or is shown in reruns. Right now, all of those graphics you see on TV are pretty much baked into the video. Suranga said it would basically require new TVs with powerful chips and Internet connections. The computer chips alone would add about 50 percent to the price of most TVs, so it will still be a while before that happens. The other option, of course, is set-top boxes. But the cable companies don’t have any incentive to allow ads they don’t control to be seen on their set-top boxes (which is why they are trying to figure this out themselves).

On the other side of the equation is placing the ads. What is needed is something like Spot Runner’s system, which lets businesses create ads and plan media buys across cable and network TV. These ads are targetable by demographics down to the neighborhood level. Google also has its own experiments with regular TV ads through a trial on the Dish Network, where it has software on Dish set-top boxes. But Google could be doing a lot more. Says Grouf

Google is not selling targeted ads now on TV. It is selling national ads through the smallest company in the satellite space. We expect them to become more aggressive, but have not seen it yet.

These two conversations keep ringing through my head. I can see a day not too far where ads on TV start to look like the text and graphical overlays we are beginning to see with YouTube, AdHoc, VideoEgg and others. But you also need an automated placement and creation system like Spot Runner’s, which today only deals with regular 30-second TV spots. Combine the two together, and there’s the future of TV.

Advertisement

Comments rss icon

  • The day this happens is the day I no longer watch TV. Period.

    I do not want crap constantly popping up overlaying what I am watching. It would be too intrusive.

    Oh, and nothing dealing with ads will always be used judiciously. It will keep getting worse and worse until we are watching our TV shows through a layer of ads covering the entire screen.

  • Interesting breakdown. You’re probably right about a lot of this, especially ad purchasing moving more and more towards the search ad model.

    I am surprised to learn that those graphic overlays on TV are “baked in” to the video. I feel like I never see those same overlays on a re-run. Most of those graphics are advertising another show that will premiere at a certain time, so a re-run three weeks later would never use the same overlay. Plus, they are never on the DVD. Is it really that difficult to separate the two?

  • I actually don’t mind ads as long as I know they are 15 or 30sec long and that they will come in certain spots on a streaming vid like what Hulu.com and VEOH.com does. Having ads wrapped around the program I think is a bit more intrusive and doesn’t give the viewer, like me, a choice to walk away.

  • “imagine a Nike logo popping up when you are watching basketball”

    Why? No need to imagine that.

    The trouble is, greater than half of TV is *already advertising*! It’s the COST of PRODUCING television programs which needs come down; there’s no reason the programming needs to cost so much. Ditch the expense of it all — we just want the b-roll nowadays. Give it to us RAW. Then go spend the $15M you wouldn’t have thrown at Katie Couric on a better HD camera instead.

  • If I’m engaged in a story, I wouldn’t pay attention to a scrolling text ad, especially if I know they’re coming.

    To communicate advertising effectively with today’s audience, the message should be integrated into the content itself… I’m a fan of branded entertainment and the new ads Wallstrip used last week.

  • When did TC turn into an web magazine. I thought start ups were featured here.

  • i don’t know, weird way to look at it. aren’t tv ads 100x more successful than video at this point. don’t the proliferation of cable channels help with relevancy and won’t that only increase as the tv merges with the Web? seems like the question is how do video ads become as impactful as tv ads. can’t think of a single brand that’s been built from video ads.

  • @2, jbenz: It is baked in as far as the delivery to your TV goes. By that time it hits your TV all of the baking is already done. The only overlays that your TV does is the channel and volume controls.

    Most shows do not have any branding or really anything but the meat of the show when they start out at the network. It is pretty raw, but it has been processed for the times that commercials will be shown as well as all the scenes and their various transitions used in the show.

    The television networks process the video as they are sending it out to give you some of the overlays, like as in what appears on most sporting events as well as the network logo that is common nowadays.

    The cable, satellite and local stations do their own mixing when they get the television feed from the network; just their editing is usually limited to splicing in the 30 second commercials that we all have already become used to.

    You don’t usually see the exact same overlays and commercials when they do re-runs because it is a multi-step process and much of it is done each time that the show airs.

    Sorry about the length of this comment, but hopefully you found it informative.

  • No one is going to “click” or care about the ad in a middle of a movie – it will be a torture to see TV. On “live” TV like sports it might be done during reruns but not in the middle of the action.
    In addition, unlike the Internet, it is harder to know information about the audience, unless and maybe the new thing – Google TV will have a tracking mechanism like it has today on the net.

  • @8, I did find it informative. Thank you.

  • Imagine a Nike logo popping up when you are watching basketball, or cruise ship when someone in a movie mentions the Bahamas. It could be annoying, but not if used judiciously.

    You’re insane.

    I’m with the first commenter — this will make watching TV intolerable, I don’t care how it is implemented. The promotional animations the networks overlay are bad enough, but “timely”, “contextual” ads? This will ruin the TV-watching experience for a lot of people, including me, and I love television.

  • The fundamental issue is that advertising and entertainment never really went well together. They were just fair-weather friends. entertainment had eyeballs, and advertising had money. each side wanted what the other had. The very nature of advertising, a buying message, is corrosive to the suspension of disbelief so carefully cultivated by entertainment. Frankly the only area where advertising (branding really) is valuable in entertainment is product placement, and only in so much as it assists the suspension of disbelief. It is disconcerting to see the protagonist pop open an can of beer that is branded “Beer”.

  • Who wants to read text when they are watching TV?

    I wouldn’t strain my eyes to read some bogus advertisement when I am relaxing/zoning out on my couch 10ft away from my TV…

  • Make product placement as easy and relevant as online text ads.

    You know how when you watch a baseball game an add is greenscreened behind the batter?

    What if this ad was a local add that related to your demographics?

    What if this greenscreen was the billboard in a chase sequence on CSI?

    The merging of videogame style advertising in TV could be the future…

  • replace add with ad :) lol

  • This is complete bullshit, they should care about the customer!

    who the hell wants ads shoved at their faces?

  • #14, Zach: I do agree I see this happening more and more and personally I think that it is the way it should be. Not necessarily by putting a moving image behind a batter, but sponsored non-moving items in the background are good.

    Some moving things may work, like if a character in a show is standing in front of a TV, why not have an ad for Coke showing on it? When it makes sense, they should do it. But if it doesn’t make sense and distracts the user too much, get rid of it.

    Television advertising really needs to go from interruption marketing to permission marketing to become really effective. If they must continue interrupting me though, then they need to find a way to tailor ads to my desires and find a way to present the ads to me in a non-distractive way, like actually placed in scenes where I can notice them. Once they do that they will have a workable solution.

  • I’ve heard that Spotrunner is anything but automated. I’ve heard they actually mail video ads to TV stations in lots of cases. Did you ever happen to ask about their revenues? I’m not sure things are quite working all that well over there.

  • Hey wait, isn’t TV dead anyway? Doesn’t the tv ad die with it? Aren’t you all going to be watching youtube in HD on your plasma? Either way, I can’t wait till all programming looks like bloomberg tv.

    @4: True. It’s the absurd cost of television production that causes people to pay $50-$100 a month to see programming that’s peppered with ads every 2 minutes, each of which costs $60-300k per 30-second slot.

    @7: False. there’s no way to measure the effectiveness of a tv ad. No GA for TV yet.

  • TV ad model is based on scale and demographics which (especially in the case of demographics) are somewhat harder to come by online. They know 18-34 year old women watch “Gossip Girl” and “Grey’s Anatomy”, and certainly in the case of Grey’s, there is enough scale to get the advertisers salivating.

    Especially with numerous shows, on numerous cable channels, targeting based on demographics is fairly easy. I understand why talking to Suranga and Nick would get you thinking about this but it’s not clear to me you’re solving a problem that anyone is aware they actually have (and I’m not sure the problem actually does exist).

    There will be more overlays for sure, but the 30 second spot isn’t going the way of the dinosaur anytime soon. It seems like those ads (the 30 second spots) should be targeted based on the market the advertisers want to reach and need not necessarily be relevant to the content (in terms of the 30 second spot). I have no doubt the overlay will grow and include things like “Sweater by XYZ designer, visit http://www.etc to find out more” Though I don’t know I necessarily view that as progress!

  • So using the speach recognition system means when Bruce Willis says the word “c#$%sucker” in a movie a text ad for a porn site will pop up? :) thats going to be great!

    Heres an idea… how about making better ads? I’m sure most of us have at least one ads that we get a giggle out of every time we see it, if the advertisers made ads that were entertaining then I wouldnt mind watching them so much..

  • @Robert Seidman: You’re right that for a big advertiser, blinkx and Spot Runner don’t provide much value. However, their primary customer base isn’t the big advertisers – those guys already have ad agencies who do everything for them. Blinkx and SR are targeting the small, local businesses who would benefit from TV/cable advertising if it could be done effectively and at low cost.

    In their ideal scenario, a basketball game is on and the video overlay or the 30-second commercial spot is for the town sports store. Or during some ABC drama, the new dentist in town can run an overlay advertising his business.

    When you compare the effectiveness of a prime TV spot to the usual types of local advertising – billboards, junk mail, flyers, odd cable TV stations – the value is pretty apparent. If blinkx and SR can do it for a few hundred bucks per spot/overlay, I think the value proposition is very attractive.

  • Here in the UK at least broadcasting regulations do not currently permit this type of advertising. It basically has to be in a continuous block, separated from programming and obvious to the consumer that it is indeed advertising. There are of course a lot more regulations :)

  • JAKE i agree if they just made better ads i wouldnt mind watching it at all if it entertains me

  • Worst idea ever. Popup ad overlays on TV? Already full of promos for other shows. I agree with the commenter who said that’s the day he stops watching. Monetizing web videos the same way will kill that too. For all the crowing about net superior to mainstream media, the web marketers are 10X worse. Say goodbye to the internet.

  • @Patrick: the primary reason I think overlay will take off is DVRs. You can fast forward or skip through commercials altogether, but if the ads are in the content, you can’t. That said, I don’t see the local auto dealership or dentist ever being overlayed in an ABC primetime drama – I think that’s far more likely for things like local sporting events on your local Fox Sports Net or equivalent.

    In primetime, I think the overlays will mostly be for the typical advertiser already willing to pay for the 30 second spot. But a.) I could be wrong, and b.) even if I’m write, there’s a lot of day parts other than primetime, and a lot of opportunity on channels besides the major broadcast networks. I’m warming up to Erick’s thinking…

  • Erick,

    Entrepreneurs and others should do well to remember the history of interactive television before trying their hand at it. I’ve spent years working with cable, telcos, and media companies and I’m a student and life long web fanatic. I’ve been pitched many times about interactive TV and ads and spent a while doing due diligence on a concept myself.

    Here’s some advice and some history:

    A Little History

    Interactive TV was invented in the 90s through a joint venture by SGI, Time Warner, and AT&T (read The New New Thing by Michael Lewis). It was a great technology feat at the time but failed miserably not even making it to market…why?…because no one wanted it.

    TV is not about “Clicks”

    Most people come home from work, turn on their TV, and turn off their brains. The web is a very different medium than a TV in terms of interaction. When people are online they get their info or entertainment in bits and pieces, the go from site to site quickly to get their fill. They read a bit on one site, do a search, and read more somewhere else.

    This is drastically different user behavior than when watching TV. Sure, people click around to find something they want but when they find it they stay and watch. Their intent is to stay not need to be constantly engaged interactively.

    What’s the Business Case?

    To date, most network investment is in making the content flow downstream faster, crisper, farther, better. To make interactive ads viable at scale, investment needs to be made for upstream data flow. True, the internet can hookup to the back of your TV or set top but that will come at a bandwidth cost incurred and passed on to the consumer. Will people pay for more bandwidth so that they can interact with ads on TV? I’m thinking they’re most likely to stick with their “internet high speed lite” or god help us dial up services to browse the web at a reasonable rate.

    Having worked extensively with cable and telcos I’d be hard pressed to see the execs sign off on a project that requires heavy investment, lots of work, no proven need or customer demand.

    Companies Already in the Game

    A few companies are already in the TV interactive ad game. Ensequence has a compelling solution, and there was a startup that allowed you to buy items seen on Desperate Housewives simply by clicking on the items on TV. A catalog would open up on screen and you could place your order…pretty cool stuff but I can’t remember their name (AgentBuddy?), maybe a clue to their success :)

    Bottom Line

    The bottom line is that the web is different than TV but one thing is always true: People will use a service if it makes things better and more enjoyable.

    There are some great web services that could be brought to TV. How about being able to search and retrieve all comedies targetted to ages 25-35 instead of clicking through 300 channels? That would make my TV experience easier and better.

    Just like on the web, make things easier, better, and faster. Interactive TV has been in play for 10 years and still doesn’t fit the bill.

    Erick – I find your posts some of the most informative and well thought out at TechCrunch. Would be great to continue the discussion with you Erick or anyone else interested.

    All the best to everyone working on making their dreams come true. Don’t you just love technology :)

    Thanks,
    Jaafer Haidar
    http://digitalu.wordpress.com

  • Lets face it, the market share of TV is sliding, and this could be the very last straw….nobody enjoys ads, hence the issue of forwarding through them or treating them as a quick snack break. If these things now start popping up uninvited it is going to be a major league irritation. Granted it will be relevant, etc, but still annoying, nevertheless.

  • How to save the horse & carriage business – strap engines on horse. uhm, i don’t think so.

  • I agree Joel – I cannot imagine anything more annoying than having ads pop up when I’m watching videos… I leave most videos if they 30 second ads placed before the video I want to watch. Those minutes add up to wasted time in life.

  • Why this constant obsession with overlays Eric? TV – like online video – needs an advertising medium that works, i.e. delivers value to the advertiser who is paying for it. Similar to your anti pre-roll stance, you seem to forget that publishers and broadcasters need to make money.

  • I Am Not Posting To Spam My Blog - March 27th, 2008 at 4:16 am PDT

    To pick up on #1: the day this happens is not the day I stop watching TV. The day this happens is the day I give up entirely on getting TV from legit sources. Since it looks as if broadcasters are finally getting round to a model in which I can watch what I want when I want, that would be a huge step backwards.

    Fact is I’m in the minority and so are the rest of us. To make a wide but probably accurate generalisation, TechCrunch readers are not the type who flop down in front of the TV and gaze goggle-eyed at whatever happens to be on for hours on end. I only watch scheduled programming if there’s something I actively want to watch at a suitable time, which on average is less than an hour a night, and I’m guessing many TechCrunch readers are the same. People who don’t particularly care about what they’re watching are more likely to accept intrusive ads, and I have a nasty feeling they make up the majority. And what’s more, they’re probably the people who are actually influenced by TV ads as well. Those of us who switch off in disgust aren’t actually generating revenue for the TV stations anyway.

  • How about whenever some tagged content is available on screen, a simple icon appears in the corner. At any time during the broadcast, you can press a button on your remote for more information.

    I picture a scenario where, for instance, you are watching Random Primetime Drama Show and a character walks on screen wearing an advertiser’s brand of clothing. A little clothing icon appears in the corner. You press a button on your remote, the picture size of the show shrinks a little bit and accommodates a simple text link of sorts saying “5% Instant Rebate of BrandX Women’s Sweaters, only at Macy’s” and you have the opportunity to cancel and go back, or click a different button and have a coupon emailed or texted to you.

    Ads are relevent and realtime, but don’t distract from the show. Advertisers would be sacrificing raw views, but the quality of actual interactions I believe would skyrocket.

    I think these, coupled with traditional commercial breaks (although maybe having 2 45-second commercials and a 15-second commercial instead of 4 30-second commercials) would be a boon.

  • @33 – that’s exactly what the start-up I mentioned was doing. Many others have dreams of this including OpenTV.

  • Jaafer, I think my suggestion is a little different than you mentioned in the fact that the whole shopping functionality isn’t built-in, only a text link that will send information to you.

    The benefit of this, as I see it, is that your TV experience is largely uninterrupted.

    I’d much rather be sent a coupon to be reviewed later, than stop the show I’m watching and buy an item right there.

  • Erick: I am a little surprised at how much time you spent writing about contextual matching and demographics vs. the real opportunity in behavioral targeting. The power of the web is about understanding multiple interests a consumer has at any one moment in time. If search taught us anything in the last few years, it taught us that we no longer need to use coarse proxies for interest (like the video that is running or demographics) but real explicit interest as demonstrated by multiple concurrent actions. Both of the examples you point out, could be small improvements, but BT (which we are doing now at veoh-plus the other stuff) is where real relevance will mark its day.

  • Hi Michael,

    I fully agree on point #1: automating the process. But I disagree on point #2; pop-ups and overlays… They don’t work in any medium (unless placed passively and activated manually by the user).

    The future of TV is all about the silent shift away from generic broadcasts in favor of the Video On Demand model combined with new platforms that automate the process of delivering highly targeted ads.

    http://www.jeff...et-to-boom.html

  • A lot of great comments on this thread. Just to clarify, these ads don’t have to be clickable, they just need to be relevant. I agree with you, Jaafer @27, that interactive TV was a failure. But that was before Web video. Whatever ad unit starts to work on Web video should be ported over to regular TV is my point. Because consumers will be used to that type of advertising on the Web.

    What that ad unit is, I don’t know. It doesn’t have to be a text link. It can be a graphic overlay. Audiences are getting used to these things and won’t necessarily think they are intrusive if they are relevant.

    As to Steve’s point @36, I didn’t mean to imply that behavioral targeting couldn’t be part of the mix. (I actually ran out of time and had to post because I had some execs waiting to show me a product demo, so I couldn’t cover everything).

    But how exactly would you do behavioral targeting on TV without tracking every show somebody watches? That raises way too many privacy concerns (although nobody seems to mind when it happens on the Web). At least with contextual ads you are targeting only based on the content within the video itself.

  • Maybe because this is coming from TechCrunch or maybe because Erick doesn’t really watch TV but what is being completely overlooked in this piece is the viewer.

    Overlay ads suck for the viewer. Constantly having the flow of your show being interrupted by what amounts to pop-up ads sounds like the thing that would finally get me to turn of my TV for good. I mean, those little station bugs are bad enough, and when they’re animated it’s downright infuriating.

  • @18 – Correct, in most cases Spot Runner sends Beta tapes to the cable operators and TV Stations. There are markets that use Fast Channel (ftp upload) and that helps, but in general SP is stuck mailing out tapes. As far as revenue, I hate to say it but word on the street is that they are struggling. Not only do they experience a high burn rate but they also don’t bring in as much as expected. ….that’s at least what I hear through the grape vine.

  • There’s been a lot of talk over the past week about the different levels of attention that the commercials on TV shows that are broadcast on television and those streamed online are getting.

    The debate started around the cancellation of Jericho and the issue of big TV networks not really doing enough to make their online audiences count in the greater scheme of things, especially in light of the fact that the shorter ads shown online don’t really allow you to move away from the screen without risking missing something while, with those on television, you can walk the friggin’ dog, cook lunch and be back in time for the next scene.

    Okay, I’m exaggerating but you get the gist. Anyway, thanks for the article, it’s definitely an interesting read.

  • I THINK YOU ARE RIGHT TECHCRUNCH

    ADMIN
    http://freeindiantv4ever.com

  • Michael and all others concerned with Spot Runner struggle. Not sure who you’re listening to on the grapevine, but have you guys seen who’s investing in and trying to invest in this company. From a media standpoint – it’s nothing less than stellar.

    This company is morphing quickly and positioning itself very well. You may be hearing struggling times for companies like Spot Runner – but the entire media advertising world is hurting right now. Economy is down, stupid companies cut their advertising as the first sense of expense reduction, when it should be darn near the last…….I wouldn’t count these guys out just yet!

  • @38

    Erick,

    I see your point and agree that the same audience used to watching video online may even come to expect the same services through television.

    The problem is, although web video viewership continues to increase, it’s still a relatively small portion of the TV viewing population and therefore it’s a hard sell to convince a cable operator, networks, etc. that it’s a worthwhile investment vs. other initiatives on the table.

    Still, here’s what might happen to make it a reality:

    - A great web video ad services comes to the world. People love it so much that they come to expect all online video to have these ads

    - Online video viewing continues to increase and the demographic is of high value

    - TV world notices and says…”So many people want this! Let’s invest in bringing it to TV!”

    - They invest, we get web video type ads on TV. It’s an opt-in service that it tested to customers with an existing high speed internet service from their provider

    Depending on how it goes, the service stays around or fails and someone 10 years from now is reminding people about it in a blog post :)

    Thanks,
    Jaafer
    http://digitalu.wordpress.com
    http://www.sensidea.com

  • i do object strongly on using people who have died for your adverts,edith piaf for one,do you not care for family feelings in using these adverts.zena

  • When is someone going to come up with a pop-up blocker for TV! Yes!

  • Chandra Xavier Raye - February 23rd, 2009 at 8:20 pm PST

    I believe CNN Money said recently that Spot Runner is going under. They’re losing clients to dotcom CheapTVSpots. Google looks like it’s having trouble, too. Ad overlays and template ads are a fad. Automating the TV media buying process actually INCREASES the cost of the media buy. Both Google and Spot Runner have shown that to be true of their systems. Besides, while I’m watching a show (online or on TV), I don’t want to stop and buy shoes. It’s insulting, demeaning, and ruins the entertainment experience.

Leave Comment

Commenting Options

Enter your personal information to the left, or sign in with your Facebook account by clicking the button below.

Alternatively, you can create an avatar that will appear whenever you leave a comment on a Gravatar-enabled blog.

Trackback URL
bugbugbugbug
Techcrunch on Facebook