Last fall and early this year I’ve been exposed to an interesting beta test at Google involving the insertion of television ads into the Dish satellite network using Google’s auction model and browser interface. The results have been very interesting, but the process carries some ramifications for an entire ecosystem of agencies, media planners, clients and broadcasters.
Let me digress right at the start by confessing a classic interactive bigotry against television as an expensive, dumb (in the sense of being one-way) medium that is dead and gone thanks to Tivo and the DVR revolution. The obituary writer with the biggest impact on my thinking was Joseph Jaffe in his polemical Life After the 30 Second Spot.
“The 30-second spot – at least as it exists today – is either dead, dying, or has outlived its usefulness. Take your pick.”
When he wrote the book (which I strongly recommend as one of the best books about the New Media Avalanche) there was no GoogleTV. What has happened is the television ad, as it existed when Jaffe stuck a fork in its butt, has morphed into something funkier and more complex because of several key changes. Those are:
- Buying. Upfronts, that practice of advertisers committing to a series or specific duration of slots is about to gets seriously messed up by Google’s auction model. Media buying is getting very efficient, (or as they say in Boston, “wicked” efficient) thanks to a model that permits a buyer to select a day, time, specific channel, specific show, and bid for the time. The winner doesn’t pay the winning price, but the price of the bid of the second-place loser. It cuts out the traditional media buying agency by putting the client right in the game.
- Flexibility: Art cards (I love Madison Avenue jargon) are the messages one inserts into the tail end of an ad to customize it, usually localizing it for the franchise or dealer paying for the slot. Example. Mercedes Benz has its agency produce a really slick commercial showing the new model swerving along the California coastline on A1A. A local dealer, Fred’s Car Barn, pays to run the spot during the nightly news. At the end of the spot, when the Ritalin Announcer speedtalksthroughtheyourmileagemayvary message, up pops the Art Card, saying “Shop At Fred’s Car Barn on the Miracle Mile in Methuen” with a local phone number. Now, thanks to dynamic serving technology from companies like Visible World an advertiser can serve literally hundreds of different art cards, modifying the final call to action to fit the time of day, show, audience, etc.
- Metrics: hand grenades share something in common with television metrics: close enough is good enough. In the 90s I was convinced “interactive television” was the next big thing, and before we’d know it we’d be clicking on lamps in sitcoms and ordering them right from the Pottery Barn. Well, I don’t have a Dick Tracy Video Phone Watch either. Television was supposed to get accountable, and fall into line thanks to the ramrod discipline of Advertising 2.0 (search engine marketing, retargeting, behavioral targeting). Guess what? It’s still sloppy, but as I’ll describe later, it’s getting better.
Now what is GoogleTV? Go back a couple years. Google is public, critics are calling it a one-trick pony overly dependent on AdSense, Google’s text advertising model that puts text ads adjacent to relevant content. It’s a good model, and it kicks off an astonishing renaissance in web advertising, but Google needs to diversify. Obviously the company is going look around the media landscape and seek opportunities to extend its auction model. Television, the Moby Dick of Marketing, is a good first target and Google puts its engineers on the task. What emerges is GoogleTV and this is how it works.
GoogleTV only works (today) with Echostar’s Dish satellite network (I’m a DirectTV customer myself). Following an alpha trial in Walnut Creek, California, they went national by co-locating some servers at Echostar’s operation facility in Colorado. Those servers directly address the Dish Network’s ad insertion system and traffic the advertiser’s spots into the slots won through the auction interface. Google purchases, in its own upfront, a chunk of the Dish Network’s local inventory. Ads are divided between national and local “pods.” The network (CBS, AMC, HBO, ETC.) sells a portion of the ad slots to national advertisers, and the local affiliates sell the remainder of the slots to Fred’s Auto Barn and the local jewelers. What doesn’t get sold is filled with public service announcements and internal cross promotions (“…stay tuned for the season premier of Lost in the House…”). Television is like airplanes, it flies, whether or not the seats are filled, and those empty seats, if they can be sold right up to the time the door is closed and the plane pushes back from the jetway is money in the bank that is otherwise lost.
Television, like any medium, doesn’t sell out 100% of its inventory. That’s why ads repeat, lots of internal promos run, this is your brain on drugs …. Google takes the local inventory that doesn’t get reserved in an upfront and makes it available to any advertiser (today mostly big companies, and soon, that mass of medium and small business advertisers who wouldn’t consider running television because the barrier to entry is so high. Larry the Dentist isn’t going to engage with an advertising agency to build a 15-second spot about teeth brightening and then build a set of different art card at the back end. Buying the ad , in the old model, was a voodoo science. Not anymore. Now Dentist Larry, if he has the commercial, can approach television advertising through the same Google web interface he would use to bid for AdSense search terms like “periodontist AND Smallville.”
I just came out of a beta test with Google and I don’t want to share results, but let’s say the cost of this program significantly less than it would have cost if we did it the traditional way. We ran thousands of slots, during prime time (these were not 3 AM infomercial remainder ads), reaching millions of viewers. Google was able to measure, through its analytic integration with Dish’s set-top box infrastructure, that over 90 percent of the ads were viewed to their full 30-second conclusion.
Measurement of the impact of the program on sales at Lenovo.com and our phone system is certainly not as a precise as the discipline we apply to search engine marketing and banner, but it was a lot closer in terms of precision than the old days in television when about the only thing one could hope for was a pre- and post-campaign brand awareness survey. We saw significant (again I won’t go into specifics) increases in traffic to Lenovo.com from specific search terms. One of the two ads was for our most excellent tablet notebook computer (which I type this on). Traffic to Lenovo.com from the keyword “tablet” exploded when the television was running, offering us some correlation with the on air content.
Let me digress in conclusion on what the big picture impact of this is for the world.
- Is Google “disintermediating” the traditional advertising agency from traditional media? This is a super touchy topic so let me be blunt and say, IMHO: yes. I can, as a marketer client, buy television, radio, print … everything but guys wearing sandwich boards … through Google’s auction front end. This means a big disturbance in The Force for media planners and buyers, the “rice bowl” of big agencies.
- Will clients need to have their own internal media planners? Yes.
- Data processing and analytics are the biggest differentiators between Google and the other global interactive ad networks who don’t own a metric capability of their own. In the future we’ll be able to make an even crisper revenue attribution and expense-to-revenue case than we did in the beta.
- What becomes of the traditional agency model? It fragments into specialty niches and the traditional agencies become high added value creative production houses. The one thing a client and a network can never do is create compelling content. Period.
The next 24 months are going to see some interesting moves by companies in the tech commerce space towards an integrated on and offline media model presented to them through a unified Google front end. The impact will be greatest down the taper of the long tail – big advertisers are moving too many dollars in bulk to care about nickel and dime economies at the auction level – but when the local ad pod – be it in newspaper, billboards, radio, or tv is integrated with search terms and banners ….
We might finally be realizing some of the sci-fi predictions made about the future of advertising a decade ago.
Fortune has an excellent interview with Irwin Gotlieb, the godfather of media planning software, and CEO of WPP’s GroupM. Here’s his take on Google:
“And then there’s the other elephant in the room: Google (GOOG, Fortune 500). If Gotlieb’s strategy is like renovating a rickety house of media, Google is trying to tear the old relic down. The search engine behemoth talks about partnerships even while launching products like GoogleTV Ads, an online auction system for selling cable ads. Who needs media buyers when you can secure commercial time the same way you buy banner ads?
“Of course, GroupM and Google do work together. GroupM spent more than $300 million last year buying search ads from Google. But that just makes the relationship complicated at best. Google is dropping $3.1 billion to acquire DoubleClick, a pioneer in selling advertisers clickstream information. (In May, WPP spent $649 million to buy 24/7 Real Media, which collects web data in nearly the same way.) Google’s heft can obviate Gotlieb’s advantages. GroupM’s clients pay in part for Gotlieb and Scanzoni to get the best deals and the sexiest ad placements. Such relationships become less important when technology makes the whole process transparent. So Gotlieb is wary. “Google can do something without regard for whether they can actually make money on it,” he says. “That can be very destabilizing to the rest of the business.”
“Google was able to measure, through its analytic integration with Dish’s set-top box infrastructure, that over 90 percent of the ads were viewed to their full 30-second conclusion.”
That’s the core – not just ability to measure but the return to a simple matter of really knowing what one pays for. The measurement will be refined quickly (at somewhere between Moore’s and Kurzweil’s speed?). But once the tv is 100% digital, no need for statistical extrapolations of nielsens.
That might look like a “media dry goods store.” Popular shows will probably still command a chunk, but the new work will be the buying not of the shows we know, but rather the buying of populations that watch ads on shows that may not be in the top whatever… Digging even out into the outlier end to find value ratios that make sense.
Will be interesting to see what this does to content, buying, agencies, etc. and what new set of marketing challenges will come.
I think one new agency challenge will be that of re-orienting targeting. “Confirmation of delivery” should put more pressure on the research, assumptions, and models that led to a given target or estimation of that target’s response.
you said: “Well, I don’t have a Dick Tracy Video Phone Watch either.”
David. I do. I got mine from Microsoft at Demo. For the record. it’s a piece of shit.
you win the “Best business post of 2008: for this well written blog entry.
Jim