The Death Of Online Advertising Is Greatly Exaggerated
from the precision-targeting dept
Over at his personal blog, occasional Techdirt contributor Tom Lee weighs in on an interesting discussion going on around the blogosphere about who, if anyone, is to blame for the precipitous decline of the newspaper business. My sympathies are with the pessimists: in principle, there are a lot of things newspapers could have done to better manage the transition to Internet-based news, but as a practical matter it's really difficult for large organizations to adapt to disruptive technologies. Tom makes some sensible points about the newspaper business, but then makes a claim about the broader advertising industry that I didn't agree with. Tom suggests that the online advertising market may be fundamentally doomed because now that advertisers can more precisely measure the effects of advertising, they're discovering that it "just doesn't work very well."
I think there are a couple of problems with this. In the first place, advertising has never "worked very well" in the sense that any given ad impression doesn't exactly get the viewer to run out and purchase the product being advertised. In the traditional advertising business, companies didn't know which specific ad will work on which specific viewer, so they adopted a scattershot approach where they exposed millions of customers to dozens of ads and hoped a few of them would have the desired effect. But despite our ignorance about precisely which ads "work" on which viewers, it's pretty clear that advertising "worked" in the aggregate. McDonalds and Coca Cola clearly get some value from the millions of dollars they spend on TV and print ads.
On the Internet, the scattershot approach is no longer necessary. Digital media allows advertisers to be a lot more specific about the users they want to target and to collect a lot more data about their effectiveness. Tom suggests that this is a bad thing because once companies discover their ads aren't working well, they'll stop spending money on them. But the flip side is that advertisers can measure when a particular ad is working, and that ad inventory becomes correspondingly more valuable. Even better, better measurement means that the average ad should improve over time. Ads that don't work can get dropped more quickly, and the ones that perform well can be put on heavy rotation, emulated by other advertisers, and so forth. That can only be good for ad revenues.
Tom also suggests that advertising is doomed because the Internet makes it a lot easier to avoid it. But peoples' hatred for advertising isn't inevitable. It's a consequence of the limitations of 20th century media technologies that required advertisers to adopt "scattershot" approaches to advertising. There was no way to target car ads at the 5 percent of the population that's in the market for a car at any given time, so the other 95 percent of us had to sit through endless car commercials. But online there are lots of ways to more narrowly target ads at people who are likely to be interested in them. In the long run, as we've said before, advertisers are going to have to realize that content is advertising. If you can make ads relevant, interesting, or entertaining, people aren't going to try as hard to avoid them. Search engines do this by only showing ads relevant to the particular keyword a user entered. Other advertisers have figured out that if they make their commercials fun to watch, people will be more willing to watch them. Of course, it's hard to predict whether the total amount of advertising revenue will go up or down over the next decade. But as long as people buy stuff, companies will be willing to spend significant amounts of money to influence their decisions.