SEO

Google’s Behavioral Targeting Flaws Put Advertisers at Risk

For a while now, Google has been testing behavioral targeting in their search result pages. And while behavioral targeting may be the next big thing, it’s already putting AdWords advertisers at risk by costing them money and decreasing their quality score.

I’ve opted to use video here to help illustrate the progression and assumptions Google tries to make. (I know it’s not a high quality video, but I hope you can see the examples used and how this can impact you as both an advertiser and search engine user)

Clearly, advertisers are put at risk with their ads being served up on irrelevant result pages. To make matters worse, it’s clear that the maximum cost per click associated with an ad can allow it to effectively dominate keyword markets it has no business being in.

Make Google Accountable

If you have an AdWords campaign, do your research and make sure you are covered here. Google’s AdWords system is something that you should use to grow your business effectively. Wasting money on poorly targeted visitors is bad enough, but with behavioral targeting your quality score takes a hit too as this waters down your click through rates.

What’s Your Experience?

Please, sound off using the comments feature below. Have you been footing the bill of bad behavioral targeting? Are your ads suffering from “poor” or low quality ratings because of Google?

Please share your experience here with other Search Engine Journal readers. A big thank you goes out to Paul Mancini at Search Marketing Corporation for passing along examples of behavioral targeting in action here.

You Might Also Like

Comments are closed.

19 thoughts on “Google’s Behavioral Targeting Flaws Put Advertisers at Risk

  1. really interesting video. fantastic find… this can negatively effect advertisers… reduction in click through rate, lower ad position, higher cost per click. just doesn’t seem like a very efficient way to deploy behavioral targeting – especially on adwords.

  2. jjisles — spot on comment. That’s why I’m blown away! We’re talking about a market leader here in Google, and the most popular system out there (and, most expensive). It’s unfair to advertisers and Google *needs* to address this.

  3. It’s a nice point Mr. Eric to raise. Alas, all my attempts to recreate your video failed. I am in Kuwait and no ads were shown at all to me.

    So here is another thing that I tried:

    1- Search for Pizza: Some food ordering ad from Kuwait shows up
    2- Search for financial report: Lots of financial ad report shows including “Pizza Sector Report”. Now I doubt this would show up if I wasn’t searching for pizza in the past. But is it relevant? First, I was trying to buy a pizza probably. Then, after ensuring my dinner is on it’s way I was searching financial reports.

    I tried sending the query again and again from a different browser (to ensure new google session) and no Pizza reports showed up.

  4. Thanks for bringing this up. It’s interesting how bad these results can get.

    My example:

    1) First search was for Toronto Dodge Dealer
    The result brought a bunch of car dealerships which is fine.

    2) The 2nd search was for Toronto Lawn Care
    The top 6 ads were all dealerships advertising their websites.

    3) Toronto Animal Shelter
    Once again the top ads in this result were all dealerships and 1 ad within those ads was actually relevant.

    Very interesting. I suggest until Google fixes this issue that you make sure you do not allow Google use your keywords like this. If you place your keywords into [keyword] they will exclude you out of these results.

  5. Behaviour based targeted is nonsense and is not going to stay for long. You can never judge a human behavior by the websites he visits cause our minds change constantly. I might be a video game addict for a couple of weeks and then switch to golfing. Howz google going to figure that out? Behaviour based targeting is easier said than done and even when it is done it is dumb.

  6. Nice find Eric. I wonder of those ads are showing because they’re set to broad match and changing to an exact or phrase match would keep the ads from showing.

    What Google should probably do is offer people the option of having the ads displayed for a behavioral match instead of a relevancy match.

    In the example with the car dealers it’s probably not so bad, but it does break down pretty fast as you found.

  7. I have noticed this in the past.

    Logically speaking I think for a given session, a person would search for a similar things. In the example itself: Dodge, Ford, and Chevy are “vehicles”. I know as an advertiser, if I were Ford, and someone is searching for Dodge, I might want to get their attention. Hence, behavioral ad marketing system in use may be doing its job.

    When it comes to curtain and lawn care, there does seem to be a flaw. I believe Google is assuming that in a given session, meaning when the browser is up and running, people would do similar searches and turn their browser off after finding what they need.

  8. Rabin, I would like to think that as well — but we’ve been recently reminded of how poor the search engines match user queries. Check this out for example:

    http://searchengineland.com/071023-093541.php

    If users are that frustrated with search results — one could assume then that they’re not leaving Google… In fact, the majority are there trying to either perfect a search query to get the results they want — OR — are abandoning the original search and giving up to search for something new.

  9. really interesting video. fantastic find… this can negatively effect advertisers… reduction in click through rate, lower ad position, higher cost per click. just doesn’t seem like a very efficient way to deploy behavioral targeting – especially on adwords.

  10. I didn’t notice any falls in revenue until January 1, when suddenly the floor fell out. I didn’t even learn about behavioral marketing until today. My suspicion is that they’re using it to displace what would otherwise be remnant or at least very poorly performing ads.

  11. First, i know this is an old post, but figured i'd comment anyways. You're missing the point in my opinion. It may seem like a flaw when you analyze it that closely, only paying attention to the search relevancy factor… However, what you're paying google for with adwords, is an advertising spot to “targeted users”, notice i said “users”, not “searches”. If i just searched for tampa dodge dealerships three searches ago, i may still be looking even though my search has changed, aka the ad is still targeted at me, the user. One thing you're missing with this example is the time factor. I'm sure google only uses this approach when the searches are close together, and as such, your search 3 minutes ago may still have relevant interest in your at the time of the targeted ad display. Thus you may still get targeted traffic from your ads.

  12. First, i know this is an old post, but figured i'd comment anyways. You're missing the point in my opinion. It may seem like a flaw when you analyze it that closely, only paying attention to the search relevancy factor… However, what you're paying google for with adwords, is an advertising spot to “targeted users”, notice i said “users”, not “searches”. If i just searched for tampa dodge dealerships three searches ago, i may still be looking even though my search has changed, aka the ad is still targeted at me, the user. One thing you're missing with this example is the time factor. I'm sure google only uses this approach when the searches are close together, and as such, your search 3 minutes ago may still have relevant interest in your at the time of the targeted ad display. Thus you may still get targeted traffic from your ads.