A Case Study Showing How Bing is Inferior to Google

by Paul Wilson

 

We have seen in the past few years Bing growing leaps and bounds in the search market. However, the growth really isn’t coming from Google’s market share, though it has nibbled some away. Instead Bing is growing because it is cannibalizing its partner Yahoo!

Growth is growth though, and Bing is becoming a significant player in the search industry. As a search marketer this is obviously very important to me. I personally don’t care how my users find my website (Google, Bing, or their mother), just that they do find it.

Yet, there is a significant problem with Bing that continues to hurt them—quality. Hate or love Google, it is hard to deny that their quality isn’t light years ahead of Bing. Due to this fact, Google recently has become extremely sensitive about Bing stealing utilizing Google results.

Search quality is actually quite easy to define. If a search engine can provide results that give you what you are looking for than the quality is considered good. Even if the search engine cannot give you exactly what you are looking for but provides original on topic websites for you to explore than again your query is considered quality.

However, if a search engine only serves results that are either off topic or completely unoriginal (all results returned are identical) than this search is considered to be of poor quality. The more poor results offered the harder it is for the user to justify additional searches.

As I use both Google and Bing I am learning that Bing falls in the category of a poor quality search engine. To prove this point I offer the following case study.*

If you type in “cheap auto insurance” into Bing you are going to find something rather interesting in the first 30 results. Eleven out of the thirty results are from identical websites. Ten out of the eleven domains are unique, but I show below that they are all basically the same site duplicated over and over, and with little to no content on the home page.

It’s clear that all ten domains belong to the same organization or individual. One indicator is that all the domains follow the same title tag structure. Obviously, the title tags are each spun differently, but if you look at the screen shot below you will see how very similar they are.

See Enlarged Image

Even if the owner was to create unique title tags it wouldn’t be hard to quickly identify that these websites are connected. Out of the ten domains, eight of them use Design A—shown below—and the other two domains use Design B (click on the image or title to enlarge).

Bing Design A

Bing Design B

 

The rankings for these domains are:

 

DomainBing Ranking
allstatesquotes.com#8
getautoinsurers.com#11
insurersbasic.com#13
hesowsshesews.com#17
lexilloyd.com#19
insurhotline.com#20
getautoinsurers.com #21 (duplicate)
insurersbasic.com#23
thehighlandviewmotorsports.com#27
proskauerrosesucks.com#28
getcarinsur.com#29

 

When you look at the whois for these domains it is not surprising to find all of them have Whois Guard Protection blocking the domain information. What is surprising is that all the domains are fairly new. Six out of the ten domains are under a year old, and the oldest isn’t even four years old.

 

DomainCreated
allstatesquotes.com2011-04-15
getautoinsurers.com2011-03-17
insurersbasic.com2011-03-17
hesowsshesews.com2009-12-03
lexilloyd.com2008-03-10
insurhotline.com2011-04-15
insurersbasic.com2011-03-17
thehighlandviewmotorsports.com2008-10-17
proskauerrosesucks.com2009-12-23
getcarinsur.com2011-02-22

 

Also, looking at the whois data you see that all the domains were registered at enom.com and all of the dns records of each domain point to registrar-servers.com, which belongs to namecheap.

Furthermore, the results have IP addresses that are close together, and seven of them even share the same C class with one or two of the other domains. Even though all the domains are with the same hosting company, Namecheap is a large entity. If these domains were with different owners I think we would see a much more significant variation in the IPs.

 

DomainIP Address
allstatesquotes.com91.205.98.135
getautoinsurers.com88.80.21.26
insurersbasic.com88.80.20.226
hesowsshesews.com88.80.20.227
lexilloyd.com91.205.98.128
insurhotline.com91.205.96.198
insurersbasic.com88.80.20.226
thehighlandviewmotorsports.com88.80.13.140
proskauerrosesucks.com88.80.21.25
getcarinsur.com88.80.12.9

 

I analyzed the top 30 results in Google for the exact same keyword to see if similar patterns emerged.

 

Google RankingDomainCreatedRegistrarHostIP Address
#1safeauto.com1997-01-14godaddy.comdnsmadeeasy.com208.94.146.81
#221st.com1994-12-08markmonitor.comatt.net12.177.226.71
#3lowpricedautoinsurance.net2009-07-03enom.comcloudflare.com173.245.61.113
#4getautoinsurance.net2006-09-25godaddy.comhostgator.com174.37.241.144
#5cheapcarinsurance.net2002-08-17godaddy.comsoftlayer.com184.173.227.211
#6cheapautoinsuranceco.com2009-06-04godaddy.comgodaddy.com208.109.181.177
#7carinsuranceconcepts.com2011-02-06dotster.coma2webhosting.com66.103.254.204
#8cheapautoprotection.com2011-09-05enom.comweb-hosting.com199.188.204.9
#9cheapautoinsurance.org2002-07-27godaddy.comsoftlayer.com184.172.239.36
#10cheapestautoinsurance.net2001-07-19fabulous.comverio.com198.171.203.61
#11cheapestcarautoinsurance.com2011-08-10enom.comhawkhost.com50.22.225.239
#12insurancespecialists.com1999-01-31godaddy.comsiteground.com184.154.225.11
#13esurance.com1999-09-03netsol.commci.com205.178.144.75
#14quality-insurance-4-less.com2007-12-06tucows.comsitesell.com66.43.59.102
#15autoprotectiontips.com2011-12-19enom.comnamecheap.com94.100.29.202
#16insurdeal.net2011-11-03enom.comnamecheap.com91.205.97.120
#17carinsuranceh.com2011-08-20godaddy.comhostgator.com174.132.188.183
#18cheapautoinsur.net2011-11-23enom.comnamecheap.com88.80.21.15
#19affordableautoinsur.com2011-11-17enom.comnamecheap.com46.21.144.138
#20autoinsurancemonitor.com2010-03-07enom.comhostnine.com50.22.35.231
#21autoinsurdeals.net2012-01-12
enom.comnamecheap.com91.205.98.135
#22insurancepolicy.org2004-06-15tierra.nettierra.net173.203.104.119
#23cheap-car-insurance-tips.com2008-05-10godaddy.comhostgator.com174.132.169.5
#24echeapautoinsurance.org2011-08-06godaddy.comhostgator.com50.22.71.9
#25cheapautoinsurancetips.net2010-05-21godaddy.comhostgator.com174.120.151.157
#26autoinsurancequoteseasy.com2010-12-06godaddy.comgodaddy.com184.168.238.1
#27cheapautoinsuranceprice.com2010-04-21godaddy.comhostgator.com174.132.165.218
#28onlineautoinsurance.com1999-04-08godaddy.comgodaddy.com208.109.122.118
#29123insurancequotes.net2009-10-27godaddy.comnoc4hosts.com96.31.81.126
#30insurance.ca.govunknowninternic.netl3.com205.225.215.50

 

When analyzing these Google results a pattern similar to Bing’s top 30 results actually did become apparent. There are five domains that are registered with enom and hosted with namecheap. These five domains also have a privacy guard on them, and when you look at the designs you see the two designs found in Bing.

 

Google RankingDomainCreatedRegistrarHostIP Address
#15autoprotectiontips.com2011-12-19enom.comnamecheap.com94.100.29.202
#16insurdeal.net2011-11-03enom.comnamecheap.com91.205.97.120
#18cheapautoinsur.net2011-11-23enom.comnamecheap.com88.80.21.15
#19affordableautoinsur.com2011-11-17enom.comnamecheap.com46.21.144.138
#21autoinsurdeals.net2012-01-12
enom.comnamecheap.com91.205.98.135

 

Though, there are fundamental differences between the five Google domains and the ten Bing domains. First, all the title tags in the Google results didn’t exactly follow the structure you see with the Bing domains. There were actually two domains that diverged from the usual title structure  used.

See Enlarged Image

 

Also, the IP addresses of the domains in Google are what you would probably see from different domain owners who share the same hosting company. There are only two IPs that are even remotely close and they only share a B class in the IP address, which is not uncommon to see with completely unrelated websites.

Next, all the domains are unique in their content. None of these five Google sites share content with each other. Also, with the exception of the two websites that are also found in Bing, the other three domains have more content then the Bing results.

Finally, a glaring difference between Google’s and Bing’s domains are that none of the five Google domains share the same design. If you couldn’t look at any of the other identifying data points you wouldn’t be able to connect these sites to each other (though, you might wonder if Design A and Design D were created by the same designer, due to some similarities).

Google Design A

Google Design B

Google Design C

   

 Google Design D

 Google Design E

   

So, the conclusion from all this data is that Google can be gamed, but Bing can be spammed. The difference between a search engine being gamed and being spammed again comes back to quality. It took little effort in Bing for someone to find ten domains, two designs, a host, and unoriginal content. Yet, considerable more effort was needed with Google in order to find five domains, five designs, five unique IPs, and original content.

A false conclusion could be drawn that it takes 50% more effort to rank in Google than Bing. Mainly because the Google results had half of the bad domains than what was found in the Bing results. Yet, the true conclusion is that it takes someone to game Google a 100% more effort than it takes to spam Bing. Simply because each site must be unique in order for Google to accept it, and with Bing you can quickly duplicate the same website 10+ times over.

And even though Google couldn’t detect that the gamed domains belonging to the same owner, it could sense what was original and what offered quality (well, quality defined according to the auto insurance vertical). Obviously, Google still has a long ways to go to fight those who game their search engine, but at least it seems to be winning the war against spam.

For whatever reason, Bing chooses not to create effective algorithms which aggressively fight against abuse and spam. As long as this is the case, Bing will be inferior to Google and will continue to lose the search engine war with them and any other search engine that can provide quality search results.

 

*Special thanks to my friend James Kelley over at Sonic Boom Lab for observing the initial data for this case study.

Related Posts: