The splash of the day is that Google is now doing trials of TV ads. It seems their grand plan is to use the cable box to track what you watch to determine what sort of ads are most effective to you.
People should note that this is a stark difference from search ads where you have direct access to what the visitor wants. On TV, people might settle and watch something that doesn’t interest them. Also, there are often multiple viewers at a time. Sometimes, there’s just nothing good on. Other times, you just leave the TV on while you fall asleep. Most importantly, how does all of this factor in services like TiVo that allow time shifting? Time will tell.
So what sort of potential does this all hold? Well, initially, I was poking holes in the concept when I was trying to equate how one would automatically figure out what ads to place after a person finished watching Hannibal. But in reality, many TV shows have predominant themes. Lost and the tropics, 24 and gadgets, or American Idol and pop culture. The point is that such a system won’t be as easy as online text ads, but it’s possible.
Imagine if you were watching a show on the top 20 hottest celebrities and the ads were only about fashion items such as clothes, iPods, sunglasses, and watches. Or during a commercial break from the evening news, you see products directly related to things you just saw in the news. I mean, it has potential. For pre-filmed content such as movies and long running TV series, this sort of optimization is done by humans. But for near-live content, human ad placement can only work off of general assumptions about what the show will likely contain.
Another interesting prospect is that ads can be clustered near the highest point of relevancy within a time slot. So Dell might want to buy placement in CSI, but the ad would come up in the point when someone mentions buying a computer. Again, you can see just how difficult such video parsing would be. But if Google is serious about TV ads, there’s no other direction to go except there. Anything else can be replicated by humans, probably more efficiently and accurately too.
Of course, the coolest part is that this is all web driven. Google’s ultimately goal is to create a single interface that lets a company publicize their product across every known medium to man. As their ad services becomes more mature and unified, Google will become increasingly more attractive to “lazy” advertisers. Should be an interesting few years.
That said, what’s going on with those radio ads? It’s been a long time since they closed that radio ad company purchase and yet we have yet to see a any real impact to their bottom line. That doesn’t sound like particularly good news, does it?
