Search Engine Monkeying

These are a few of my favorite things:

• pg 76 - Relevance between the Headline and the Display URL.
• pg 90 - Long Tail keywords don’t work (indirectly related terms).
• pg 101 - Local ads help organic search results on Google.com.

There is also some interesting info on Alternative Search Engines (2nd
Tier) but the comments are totally contradicting. Did anyone else see
anything else particularly interesting? Do you think some of the
conclusions drawn from any of the experiments is incorrect?

Popularity: 83% [?]

Just a short list of techniques to lower your average cpc.  Please comment if you have more.

1. Separate content and search campaigns.  Content has low CTR. Search has a high CTR. Separating them improves your Quality Score.  Content ads should be bid at nearly half of search.
2. Separate geo-targeted campaigns.  Foreign campaigns are generally worth less then US campaigns because they have a lower ROAS. Lower foreign campaigns appropriately.
3. Remove keywords with a low quality score from an AdGroup.  Put them in their own AdGroup.
4. Remove keywords that do not display an ad when queried.
5. Group your keywords to be more relevant to your landing page.  More relevancy, higher Quality score, lower cpc.

Popularity: 51% [?]

I really like this PageRank detector for Dig (not affiliated with Digg).  The detector pings your site from a number of different services and then displays the I.P. address and the PageRank your site received from each of the sites.

Dig Pagerank in 700+ datacenters

Blogged with Flock

Tags: Googlepagerank

Popularity: 21% [?]

Email Insider has an interesting article worth checking out. The premise of the article is combining two ad copy that would appeal to the thinking of a left brain imaginative thinker and a right brain thinker who would be more focused on details. Though, their examples are meant to be applied to email subject lines I think the the creative marketer would be able to make the theory work in SERP ads.

Popularity: 9% [?]

(I’m taking Master’s classes and this is another post expanding upon a topic that came up in class).

In class the other day we were talking about search arbitrage and if the practice is inherently evil.  Arbitrage is an economic term describing the “practice of taking advantage of a price differential between two or more markets“.   Search Arbitrage takes advantage of the inherent difference between the cost driving users to your site using a search engine and the amount advertisers are willing to pay a site for the users viewing their ads.  Taking advantage of the relevance and immediacy with which you can customize a search campaign, marketers are able to target a news story that lots of people will be interested in and deliver the traffic to the publisher’s site.

There have been several articles questioning the morality of Search Arbitrage from a customer stand-point; Search Arbitrage: Good or Evil by Catherine Seda and Search Arbitrage Issues on Search Engine Roundtable.  These articles question the decline in the user experience that poor search arbitrage can lead to.  Driving traffic for the sack of driving traffic is like herding buffalo off a cliff.  Neither the landing page/site advertiser nor the user get the experience or the relevancy they desire.  The only person who seems to benefit is the publisher from inflated page impressions.

Additionally,  recent articles have criticized marketers for buying keywords associated with tragedies.  While this may be considered poor taste, I don’t feel that this type of search arbitrage detracts from the user experience and instead, adds additional value to the search engines, the publishers, and the advertisers on the landing pages.  Because of the redundancy and precautions taken by search engines, new content doesn’t have as much relevancy and essentially must “buildup steam” before it will begin to rank high in organic listings.  It’s the marketer’s responsibility to second guess ranking and historical data and determine what consumers are really looking for.  Otherwise, marketers should just let computers do their jobs.  For the Virginia Tech massacre for example, search engines would have organically returned older articles, probably related to basketball or the NCAA, with the keywords “shooting” and “massacre” until the search engine algorithms would have determined that these articles were no longer relevant to consumers.

It is definitely easier for a marketer to determine that  “Virginia Tech Massacre” or “World Trade Center disaster” (for that matter) are going to be high traffic keywords.  However, the human element that search engine marketing offers to the equation can instantly make decisions about consumer relevancy that a computer cannot.

Popularity: 9% [?]

I’ve been trying to weed out competitors using our Trademarked terms in their ad copy and I’ve been having limited success. Usually, a trademark holder would just contact the offending party and threaten to sue them. This is relatively successful and there are previous cases that set a precedent so, more then likely, you’d win. Our trademark violators, on the other hand, are distributors that simply refuse to play by the rules. Management has been slow to discipline the parties that are infringing on our trademark so eventually I decided to use Google’s trademark complaint procedure to enforce our trademarks.

The whole process took a couple of weeks for the trademark complaint to take affect but eventually Google disapproved all the competitor/distributor ads with our brands in the copy.

Thats when things got interesting.

1) The Google Trademark Complaint does not recognize variations in the display URL. Display URLs such as BuyApple.com are perfectly alright to use as well as any variation of the domain. Instead of .com, if you can find .net, .biz, or .us you can use the trademarked term in the display URL. You might not think this is a big deal but the text in the display URL is still bold.

2) Expanding on the previous topic, Sub-domains can also be used in the display URL and will be bold when the user searches for the term. If the user searches for “ipods”, and your site is ipods.mysite.com, ipods will be bold in your display URL.

3) Google requires the trademark holder to notify Google every time a variation of the trademarked term is used in the ad copy. Kudos to Google for letting trademark holders expand beyond the basic term or phrase, however, there is almost an infinate number of ways to get around the trademark term without using the trademark. Using a capital i instead of a l, adding the ® directly to the end of the phrase (or any number of special or spanish characters), adding a space to the term, or even basic misspelling. How many consumers read the whole phrase anyway especially if the phrase is a long word, or technical word, or two different words together?

4) Lastly, an advertiser can always use the dynamic keyword tool to insert the trademarked term into the ad copy dynamically.

I have to admit that there are some special cases that I think trademark cases are overblown and maybe it’s best Google shows some restraint. The recent Zale’s dating ring ruling is a perfect example of a company taking trademark protection too far. In my own experience, a company trademarked the term “Type 1″ even though there are at least three different common uses for the term (think diabetes, fonts, Levi’s Jeans).

With Google being such a powerful online entity it has to do something about trademark infringement. Maybe instead of focusing on the term, Google should focus on companies that hope to confuse consumers. Thats the reason trademark laws were written in the first place.

Popularity: 10% [?]

Google Released a new beta feature today called the the Account Snapshot that closely mimics Yahoo Panama’s Dashboard. Accessed in the top navigation tabs, the Account Snapshot contains Alerts and Status, Announcements, a help box and overall campaign performance. I have the view split into two images but the they appear as one screen in the Google account (this account just had conversion tracking added and was using third party tracking and that is why the stats are so poor).

Popularity: 10% [?]

Initially when I started this blog a couple months ago, I was surprised by all the ads that appeared on the site that had little or no relevance to the content. It took me a little while to determine that these advertisers are using the term “engine” in their AdWord content network campaigns and with very little content on my site, Google saw a relevant connection.

Eventually, as I added more content to the blog, Google’s algorithm was able to determine that my site had nothing to do with small engines and dirt bikes (I’m still not sure why the ad for www.INGDirect.com is relevant). However, if the advertiser had run the negative keyword suggestion tool in Google, they would have been able to avoid having their ads appear on a completely irrelevant site.

This probably won’t hurt the advertiser is the short run. I don’t think many of my readers are going to click on these ads and the advertisers will not have to pay from the impressions. However, the longterm effects of a low click through rate (CTR) and an apparent absence of relevance will lower the advertisers position and increase cost per click (CPC). Additionally, with content space being very limited (only 3 or 4 ads per group) the effects of not maintaining a high position can have a detrimental effect on their campaign.

 I sent an email inquiring about the campaigns on my site to each of the advertisers listed in the example.  Basically, I asked the following questions:

1. What keywords (or the ad title) do you think attributed to your ad showing on my site? 
2. Do you use a targeted or long-tail search engine strategy? 
3. Do you do your search engine marketing in-house or do you use an agency?

I’m interested to read the responses I get back. 

Popularity: 11% [?]

When people search for “Life Quotes” they are often looking for quotes about life.
Why do the major search engines only give them hundreds of pages of
commercial garbage about Life Insurance Quotes?

–Bill Austin

Popularity: 9% [?]

What do you do if you have a product that is converting poorly? Currently, a product I manage is getting a measly 600 clicks a week. This amount of traffic is just pathetic and I have been talking with the brand owner about increasing visitors to the site. However, the only reason I am reluctant to do this is that the conversion rate for this landing page is a horrible .15%. That means in a nutshell, if 10,000 visitors clicked on our ads and visited the site, only 15 people would make purchases. At this rate, even if our cost per click was a meager $.50 and our average order value (AOV) is $90 we would be losing money hand over foot.

(2 orders * $90 AOV) - (600 clicks * $.50 CPC) = -120

Extrapolating this data, if we increase traffic a ten-fold:

(20 orders * $90 AOV) – (6000 clicks * $.50 CPC) = -1200

Obviously, the landing page needs a lot of work. That is the biggest problem since I am only advertising to the core traffic  associated with this product. Additionally, I was trying to figure out the standard deviation of the conversion rate in order to justify not driving more traffic and I remembered a couple of concepts.

1. Standard deviation is irrelevant in this situation because an increase in sample size could dramatically change the average. If I make my net bigger, I may catch more fish.
2. Advertising and Marketing drive sales and not the other way around.

Obviously, what I am doing isn’t working and needs to be changed (lower Max CPC) but lowering my overall budget is not the answer. Being more creative and aggressive is.

Popularity: 9% [?]