Google is moving toward a method of displaying ad groups that are personalized based on a searcher’s previous searches. With this discovery I am going to write this article in two sections (and hopefully one part).
The first section is going to focus on Adsense targeting from the advertiser’s viewpoint. This approach, known as placement-targeted or site-targeted campaigns, allows an unprecedented level of customization and targeting by the advertiser. Not only can the advertiser target a particular website, but a particular page or a particular sub-group of pages can be targeted. We will look at how Google is allowing advertisers to pick sites from selected lists and how they set different bids for different sites.
The second section will look at how Google is remembering a searcher’s previous activity and using this information in displaying Adsense results on web sites the user visits. Now Adsense is not only dependent upon the site’s content, but also on the search behavior of the user, independent of cookies and even logging in — though the user (or the computer) must have used the search engine at a prior time.
We will also quickly discuss whether, at the end of the day, any present advert placement application can realistically catch up with Google’s Adsense in the next three to five years, or whether creators of such applications should simply seek another line of work — social networking e-commerce centers, perhaps.
Google advertisers can increase their click-through rates by creating placement (or site) targeted content campaigns, choosing to position their ads on specific sites. It’s content control/placement on steroids. By picking selected sites and bidding, you can "ensure" that your ads get seen by a much wider audience; all this definitely changes the size of the budget, but that’s why we pass on the cost of the advertising to the customer, eh?
It all started that way, and now you have site-targeted campaigns, consisting of ad groups that contain lists of Web sites, not just keywords. Placement-targeted ad groups can include all content ad types: text, static, and animated GIFs, even video with sizes ranging from skyscraper to banner to defined boxes. Advertisers can choose whether to be charged on every one thousand page impressions (CPM) or on a cost-per-click basis.
According to Google Adwords, advertisers can choose to display their ads on specific pages within a site, or even subsections of pages. For example, advertisers can choose to advertise on the New York Times Web site in the business, fashion, health, travel, or sports section.
Google’s content group is huge and growing rapidly. Some say new sites are joining at the rate of at least 100,000 per month. Information on some of the top sites in the network can be found here. Almost every major media channel is represented, from leading magazine and newspaper publishers to television networks.
The ads are then placed with the sites you choose. This is a improvement on the normal placement process, which is stated by Google as "Google continually scans the millions of pages from the content network to look for relevant matches with your keywords and other campaign data. When we find a match, your ad becomes eligible to run on that page. Google’s extensive Web search and linguistic processing technology can decipher the meaning of virtually any content network page to ensure we’re showing the most relevant ads."
Normally advertisers find which Web sites display Google AdWords content ads by using Google’s placement tool, using a list of categories and keywords, and also listing specific URLs of sites, or by setting the desired user demographics. The tool digs into the pool of site publishers and lets you choose which sites will carry your ads.
As a quick rundown of what I have described in the last few paragraphs, here is what Google has to say, in brief. (This can be found here).
You can select sites for your placement-targeted campaign in two ways:
Name the sites where you’d like to advertise.
Give us a list of words that describe the type of site you’re looking for.
A site must be part of the Google Network in order to be available for your placement-targeted campaign.
A big advantage of winning a bid to place an ad could be that your ad could occupy the first four spaces, since you could stretch your ad to occupy four spaces. This is due to the expanded text ad specification of Adsense. If you specified that a strip of four AdWords ads should appear in a banner at the bottom of a page, placement-targeted ads will muscle the other ads off the ad unit and be displayed.
This is the real peach. The results are here for users to see; just search for anything on Google, go to a site that uses Adsense, and immediately you will see the results. I can see this on my personal blog already. And seeing this makes me seriously wonder whether any search engine ad placement package can compete with Adsense — but first, let’s look at some of the original apps that used the same theme (personalized ads) but which are no longer with us.
The first website to utilize personalized ads was www.findory.com. The website is now defunct, but it was the first piece of evidence that it was possible to display Adsense adverts based on a user’s past behavior. A quote from the founder’s blog says, "This early version is built on top of Google AdSense, but these are not normal AdSense ads. They are not targeted merely to the content of the page, but to the individual behavior of each reader."
Findory worked out a system that integrated its user data with Google’s ad network to generate much tighter ad matching than Google’s page reading contextualizing algorithms. Findory passed along accurate data to Adsense about the source on the page, allowing for a variety of useful ads.
Nobody was exactly sure what Findory.com did; it may have been Adsense testing or some sort of hack. But when they disappeared, Google started personalizing ads on computers. Note the difference — before this, Google based ads on the web site’s key words. Now ads are served based on the computer’s search history.
The industry noticed personalized adverts about a year ago. This post on webmasterworld revealed when a user noticed the new personalized searches. Danny Sullivan noted that the ads were extremely intuitive, stating that they seemed not to be tied to login, but to stored data on the browser. He states that he "just searched for travel, then for sale, and bang! There are some travel ads showing up.Then I searched for computers, then online and got computer ads. I moved over to another browser (without any stored history), searched for online, got nothing. So I searched for computers, then online — boom! Hello ads, lots of them, all about computers.
Google sent this information to Search Engine Land: "What you’re seeing is that we look at the user’s previous query and see how well it intersects with the current query. If it’s significant, we’ll use it to help targeting on the current query. We simply look at what’s in the referring URL (every time you load a web page, the HTTP header includes your previous URL as the "referrer")."
The illustration Search Engine Land gave in their article as a test (you can find the article here) showed that you can do a search for weather forecast, then conduct a new search on holiday in Spain, and then do another search on weather forecast and you should then see ads for weather reports in Spain.
Disadvantages to People Targeting
According to a user on Webmaster World, the above tool has its hitches. He complains that "I think it sucks, if it’s indeed true! I came in to the forum today to post and ask about seeing tons of ads on my site lately which do not even remotely match those showing up in the Google Adsense Preview Tool and realized this is probably the same issue! I just realized at least half of the ads I was seeing related to old topics I was interested in LAST MONTH." A lot of users complain and worry that it will dilute their brand.
It seems this experiment has become an integral part of the Google ad placement application. This leads us to wonder whether this will lead to an increase in Google’s brand popularity as advertisers realize that their links are more likely to get clicked on. Publishers, too, may enjoy it as they see a massive rise on the number of click-throughs for their sites. There will be the inevitable cry of privacy advocates as they see "Big Brother" Google, spying on us all; but I am not one of them, and until they make the news, I think paid campaigns should take notice, and take advantage.