Quantcast

Irrelevant Keywords Can Be Costly

How Effective is Google's Session-Based Broad Match?

Get the WebProNews Newsletter:
[ Search]

Some Google AdWords advertisers are not pleased with what they are finding in Google’s Search Query Performance reports for their campaigns. These reports show advertisers what keyword queries are surfacing their ads, and some are finding some of these keywords questionable. 

Are you losing money on clicks from questionable keywords? Let us know.

You might think that an ad impression is an ad impression, but when you’re charged by the click, you want the clicks to come from people who are likely to buy what you’re selling, considering that you are paying Google for each click. 

A Wall Street Journal piece has put the spotlight on some of these advertisers, including a New York dentist who claims irrelevant keywords have cost him nearly $3,000 over the last year or so. The problem allegedly stems from Google’s session-based broad match feature, which shows ads to users not only for a single query, but also for subsequent queries in the users same search session. 

Google explains the feature in the AdWords Help Center:

"When determining which ads to show on a Google search result page, the AdWords system evaluates some of the user’s previous queries during their search session as well as the current search query. If the system detects a relationship, it will show ads related to these other queries, too." 

"The system considers the previous queries in order to better understand the intent of the user’s current query. The added information allows the system to deliver more relevant ads."

"This feature is an enhancement of broad match. It works by generating similar terms for each search query based on the content of the current query and, if deemed relevant, the previous queries in a user’s search session. Your ad will potentially show if one of your broad-matched keywords matches any of these similar terms."
 
Sounds good in theory, but the advertisers complaining appear to disagree with what Google is considering to be relevant. The dentist from the WSJ story cited "penis enlargement" and "[Chinese characters] in Chinatown" as examples – not exactly dentist-related. The story also cites a plastic surgeon, who counted "olivia newton john photos" among questionable keywords. 

The WSJ spoke with Google’s Nick Fox:

Nick Fox of Google Explains Session-based broad matchMr. Fox acknowledged there are "edge" cases in which search queries "does not appear to be relevant to the ads, but the context of previous queries indicated that the user would have a strong interest in that advertisers’ ad." In addition, he said, "a user must be interested enough in an ad to want to click on it." He said a very small percentage of ad clicks are session-based and that advertisers can limit the scope of their campaign to halt session-based clicks.

Google’s Mr. Fox said: "It has to be the case that the users, in the very recent history, searched for terms he’s advertising on."

It’s worth noting that Google says that whenever an ad is served based on the associated keyword’s relevance to the previous search queries, the ad’s performance has no effect on that keyword’s Quality Score.

It’s also worth noting that not everyone is unhappy with the session-based clicks. Jordan McClements, commenting on a Clixmarketing post on session-based broad match says, "If you are in a niche where there is not much search traffic, and a new client/sale is worth a lot of money to you then it is probably a good idea to keep all your ‘broad’ options open."

John Lee, who wrote that post, says, "I want advertisers to be aware that in the case of session-based broad match – you can’t turn it off. My recommendation is to remain vigilant in reporting, primarily with Search Query Reports to ensure that the session-based query matches that do come through are relevant. If they aren’t, roll that knowledge (and those queries) into your negative keyword lists."

Probably good advice. 

Perhaps the real question is how much of the problem is Google and how much is the advertiser? 

Speaking of negative keywords, Google actually just released a new feature this week to manage negative keywords across multiple campaigns with negative keyword lists. 

Have you wasted money on irrelevant session-based clicks? Comment here.

Irrelevant Keywords Can Be Costly
Top Rated White Papers and Resources
  • http://www.the-barcelona-guide.com crispy_barcelona

    Are you sure this is related just to the session function? If i remember rightly about a few years ago Google changed the definition it uses for “other relevant variations” (google broad match)
    within broad match and from then on we starting seeing more liberal “interpretations” of keywords, for example returning “service Madrid” for “service Barcelona” Interestingly the “errors” I’ve found in often match errors which the Keyword Selection tool returns too. Somewhere along the line the relevancy algorithm isn’t functioning right.

    • Chris Crum

      It’s certainly possible that advertisers are showing for non-desirable keywords for a variety of reasons. Again, though, how much of it is the fault of the advertiser’s own management of their campaigns? I don’t know the internal practices of the advertisers mentioned, but it’s something to consider (to Lee’s point referenced in the article).

  • crispy_barcelona

    Yep, there is definitely a case of “a bad workman blames his tools” behind a lot of failed Adword campaigns. However, when Google broad search transforms “Barcelona” into “Madrid” or something similar you’re stuck between a rock and a hard place. Either you go for a narrower match option, and risk losing other relevant searches, or you list major Spanish cities in your negative keywords…pretty time-consuming.

  • http://www.simplyclicks.com Simply Clicks Training

    I’ve spent years of PPC training trying to tell people to beware of broad match.

    1. If you haven’t got time to review the search query reports
    2. If you haven’t got time to build a negative keyword list
    3. If you don’t measure cost of sale or cost of acquisition

    Then you shouldn’t use broad match.

    Despite using modified broad match (that is the + version) when I need volume, exact match and phrase match almost always beats broad match on conversion metrics.

  • http://www.ZenithMotorsports.com RML

    I have paid quite a bit of my advertising dollars on irrelevant keywords. Beware and monitor closely!

  • http://www.chassis-plans.com Rackmount

    I advertise for “rackmount computer”. I’ve never advertised for “rack”. Yet I’ve gotten clicks on search phrases such as “sheep sheering rack”, “lipstick rack”, “bike rack” and so forth. Hundres of different versions of “… rack …”. Google seems to think that “rackmount” is the same as “rack” and that anything with “rack” in it must be a “rackmount computer”. I’ve got several thousand negative keywords including “boob” (suggested by a female Google rep) because the slang for female breasts is “rack” (as in she’s got a nice rack).

    People will click on anything. My ad is very specific about computers, yet somebody looking for “sheep sheering racks” clicked on the ad.

    On the other hand, Google does not extend the same extension to mispellings of negative keywords. If somebody enters a phrase that should clearly match a negative keyword but they’ve got a typo, Google will ignore the negative and show the ad. So “bik rack” will show my ad when they wanted to type in “bike rack” and I’ve got that as a negative.

    Looking at metrics, there are literally thousands of appropriate search phrases used that my ads show for so using exact match is not an option. People construct long tail searches sequentially so “rackmount’ becomes “rackmount computer” becomes “rackmount computer military rugged” when they should have started with “rugged military rackmount computer”. So I advertise for the words I think are appropriate and just keep adding to the negative keyword list.

    What is really frustrating is Google won’t admit their synonym engine is flawed. Why should they because it generates additional income for them. They do everything they can to display high price ads. An ad for “rackmount computer” is a lot pricier than for “lipstick rack”.

    [This is shameless SEO - rackmount but the above is heartfelt. I've got the email strings with Google to prove it.]

    • http://www.chassis-plans.com Rackmount

      I forgot that I actually do advertise for “rack” as a broad match at a very low cost in an effort to catch the bad synonyms. Hasn’t worked. Because of the low cost, I get very few impressions and clicks. Google would rather display a bad high priced synonym instead of a real match that has a lower CPC.

  • http://jaymassey.com Jay Massey

    Good information Chris, but is using Broad Match and not monitoring it the user’s issue or a Google issue? Nevertheless, thanks for the info.

  • http://www.electric-reviews.org Mark Demers

    I use broad match keywords on my sites home index page and then more targeted ones throughout my site.
    i don`t advertise too much but when i do i use specific keywords- i still get irrelevant clicks too. My reason for targeted is so you don`t get as many irrelevant clicks ,but when someone clicks on your ad for “rackmount computers” and their search phrase was” lipstick rack” or “bike rack” – What can you do?, this happens on Facebook too .I could be wrong but this is probably a problem on all of these types of mediums.

    Good post ,

    Thanks

    Have a Great Day !

  • http://www.known.ir aboozar hoseini

    UNKNOWN introduce to you the most of plants.
    you will know the most information about medicine plants by the shortest sentence.
    Unknown guides you how to use medicine plant.