No, this isn’t a call for revolution.
Disambiguation is the process by which paid search on the Internet tries to get smarter by understanding semantics, context and the intentions of the person doing the searching.
For example, if you type "bugs" into Google, are you looking for information on insects, Volkswagens, audio surveillance technology or annoying co-workers?
The semantic search experts are supposed to be working on this, but it appears disambiguation is still a ways off.
Take these two examples:
- While reading about the Palestinian/Israeli conflict in Gaza on a news website, I noticed a banner ad promoting a peaceful vacation in sunny Israel. While that sounds like a lovely holiday to consider, the context was distasteful while reading, as I was, about the devastating military conflict.
- While reading on another news site about the U.S. admission of torture at Guantanamo Bay, an ad showed up offering hotel discounts in Guantanamo. The ad-serving tool obviously had been programmed to read text in the article, find any geographical location mentioned, and insert it into a hyperlinked sentence about hotel discounts. Being curious, I clicked the link, which went to a generic page that then led to Expedia and several other travel sites. Not surprisingly, there were no actual hotel discounts in Guantanamo, Cuba.
This all goes to show that Internet advertising has substantial room for improvement. No doubt, someone paid for all those ads and click-throughs, and those impressions and clicks will show up on some report to validate how successful the campaign is—when, in fact, it was all money down the drain.
As with all media buying, it’s easier to throw money at something than apply intelligence to it. It reminds me of the careless television, radio and print media buying that had characterized the ad agency business for decades. That’s why mass media has always been a suspect marketing tool, while trade media advertising is often a very smart, targeted marketing buy.
It’s all about precision and having confidence that the right people are seeing the right message at the right time. That's a tall order, and the robots at Google still have a ways to go. They need to work a little harder at disambiguation. Maybe they need a few more humans.