Effective Internet Marketing Strategy and Technique Through Experiments, Measurement and Audit

Improving Upon Google AdWords Training - Part 1

You’d think a multi-billion dollar commercial organisation, raising its funds through advertising, would offer a training course that helped advertisers. You’re right! The AdWords Learning Center offers a free training course, leading to a third-party administered fee-bearing exam.

Regrettably, Google’s pursuit of wealth, or just plain incompetence, has lead them to some gilding of the truth. This series of blog articles looks at some of the errors in the current course material and the impact that they have. Most of the errors appear to be a result of ignorance within Google, or excessively enthusiastic promotion of the service. You’ll need to know Google’s “correct” answers to pass their exam, but if you are an advertiser, you should also know what actually works, and why. Here’s a blow-by-blow identification of the errors in the AdWords Learning Center training and the Quick Quiz offered.


Lesson 1a - Introduction to AdWords

The quick quiz for Basic AdWords Features contains a question about “Advertisers have the option to show their ads on _”. Google asserts that the correct answer is “the Google Network” (the third party partner sites, such as AOL and MSN that use Google’s adverts). Google incorrectly gives you no marks if you select the option “Google Search Pages” - but it is right!

To confirm this, use the Campaign Edit Settings page and deselect the options for “Google Search” and “Search Network”, leaving the “Content Network” checked. Under these conditions you can have adverts running only on Google’s Content Network (AdSense and similar non-search publisher sites) and not on Google Search Pages.

I don’t see any ulterior motive for this - it is probably just a failure to understand their own complex advertising system.

If seriously nitpicking, the wording of the first question in this quiz (”what is the clickthrough rate of an advert that receives 30 clicks for 1000 impressions) is also misleading - it assumes that all clicks and impressions are billed, rather than received. Strictly the question should be about billed clicks as Google routinely delivers more clicks and impressions than it bills.

Lesson 1a - Introduction to AdWords - Benefits of AdWords

The next quick quiz, Benefits of AdWords is also subtly flawed. The second question (”Google ads can appear within ____ minutes…”) should be “Google text ads can appear on Google Search Pages within ____ minutes”. Image adverts and access to the partners on the Google Network and Content Match is not permitted until an editorial review has been conducted - taking between one and three working days. It is only the Google search pages that can carry a text advert without editorial review, and even then, the advert must not violate editorial policy such as trademarks or the use of superlatives.

The next question “Ads appear on Google ____” is meant to establish that Google delivers adverts in response to users’ search queries. So it marks as wrong, two answers that are also correct. Google’s recent introduction of Ad Scheduling allows advertisers to choose when, during the day, adverts are eligible to be shown. Additionally, selecting a lower budget than Google recommends results in advert display rate being reduced from showing on every search to showing on a percentage of searches - randomly allocated. All three answers are correct!

The very next question, intended to show advertisers how flexible the system is, also overlooks a problem. “AdWords allows ___ changes per month” is supposed to have advertisers excitedly selecting “unlimited” as the answer - but some operations are subject to a quota. Each campaign, for example, has a limit of ten budget changes per day. Strictly speaking the delays in notifying all Google’s servers of changes and reporting the changes, means that effectively there are about 96 intervals for bidding changes per day. OK, it’s a quick quiz. But the desire to enthuse advertisers results in offering misleading questions with insufficiently thought through answers.

Lesson 1b: AdWords Policies

The quick quiz about invalid clicks contains some misleading statements probably intended to allay fears. For example, the first question (”Google filters invalid ad clicks after they are reported to your account”) is claimed to be correct if you answer “False” - but in fact Google’s invalid click detection software continue to operate, with clicks and impressions being removed from the account for hours or even days after being initially added. So Google does filter invalid clicks after they are reported.

Another answer is intended to allay fears, rather than reflect the truth. “What type of clicks does Google filter?” is intended to elicit “Clicks by automated clicking tools”. However, botnets are probably one of the most severe forms of click fraud and are tremendously difficult to detect, if properly set up. So the very last type of click that Google can actively defend itself from, is a cleverly constructed automatic clicking tool. It defends itself best from people who repeatedly click on the same advert, every time, with the same browser - for example, the classic “double clicking” problem in which users trained on Windows “double click to open”, do the same on all web page links.

More on invalid clicks

The further quick quiz suggests that “if your ad is receiving many clicks but has a low return on investment, ____” and is looking for you to say that you need to optimise the advert. My general impression is that most companies need to optimise the landing page as much as doing anything about the advert. Landing pages (Destination URL) often do not clearly show the user what to do next.

Lesson 1c: Ad Distribution

The What is the Google Network quick quiz question “image ads are only shown when ___” predates the introduction of site targeting. You can now have your adverts show “when they match the content of the web page”, but also “when you have selected a site to display your adverts on”. Admittedly Site Targeting relies on identifying sites that have relevant content, but there is a material difference between having your advert shown on any matching page, and only on selected sites.

In the Ad Distribution: Common Questions Quick Quiz the question “which of the following is true regarding the Google content network” is disingenuous or positively misleading. The answer given by Google is that it “is not necessary for advertisers to write different adverts for search and content properties”.

Content Match advertising is different in type and intent from keyword search. Consider when you see the two types of advertisement. When you search, you are actively looking for an answer. You respond to the organic and search links because those are what you wanted to see. When you are reading someone else’s blog, or a news site, watching video clips or whatever - you aren’t searching. You are browsing, and Google recognises the difference in user intent implied by that - it does not penalise you for low CTR on content match adverts. Google does know that Content Match is not the same as Keyword Search.

A well recognised idea within sales and marketing, known variously as the sales pipeline or the buying process, is a model of buyer behaviour. At the earliest stage, a prospective buyer becomes aware of the need. After that the buyer does research, then comparison, makes a decision and an acquisition and then enters post-purchase evaluation. Keyword search most closely matches the active research and decision phases. But content match most often matches needs awareness and post-purchase evaluation. In any book on Marketing Communications you’ll find that addressing these different stages requires different messages and may even require a different landing page. In practice, Merjis has found that the most effective adverts are often different on the two networks.

We believe that Google’s approach to content match is disingenuous and motivated by greed. But if you want to pass the exam, you have to pretend that this blatantly false proposition by Google, is true.

Lesson 1d: Pricing and Ranking

No objections to any of the quiz answers and content.

"Improving Upon Google AdWords Training - Part 1" was published on October 17th, 2006 and is listed in google, adwords, training.

Follow comments via the RSS Feed | Leave a comment | Trackback URL

Improving Upon Google AdWords Training - Part 1: 4 Comments

  1. Ian Feavearyear wrote,

    Great Jeremy, excellent stuff! Can’t wait for the rest!

  2. Rich wrote,

    Nice detective work, Professor Chatfield! Google should hire Merjis to audit the quizzes. Seems to be a lack of Quality Control. Whenever the AdWords system changes, they should go back and check all of the quiz answers. Clearly, that’s a big task with the frequent changes. As you’ve noted above, changes like the introduction of site-targeting change some of the answers. Same with Starter Edition. Many questions almost need a qualifier like, “Not considering site-targeted campaigns or the Starter Edition, blah blah…”

    Your comments about the content network distribution for section 1c are spot on. I haven’t looked at these quizzes in awhile. When I just read question 4 again, after having run many ad campaigns, the first answer “Advertisers should write different ads for search and content properties” did, indeed, seem like the correct one. While “It is not necessary for advertisers to write different ads for search and content properties” is true, it is not *as* true as the first answer. ;-)

  3. Chatfield Jeremy wrote,

    Hi Ian and Rich (I know these guys from the AdWords Help Forum). Thanks for the comments. I’ve been looking at the AdWords Learning Center for a course we’re offering on paid search. I’ll type up the further problems as quickly as I can! We’ve been working on some other research with content match improvement and there have been some other experiments that shed some light on the AdWords Quality Score… So I’m not sure yet what the next article will be! I think that the Content Match work offers some pretty dynamite insight.

  4. James wrote,

    Jeremy, I’m honestly convinced you know more about AdWords than the people at Google. The Adwords system is a monster that not many can completely grasp but you seem to come fairly close.

    I’ve noticed a few of the quiz errors as well.

Leave Your Comment

Is this article any good? What helped you? What made you think it was wrong? What else would you like to know or discuss?

Merjis Internet Marketing Blog is powered by WordPress and the YUI-Mainstream Theme by Buzzdroid.com