Posts

wrecking ball

What Miley Cyrus Can Teach Us About Bing & Google Image Search Results

In a previous blog post we analyzed the differences between identical searches in Google and Bing; however, we never compared image search results. They are probably pretty similar right?

Actually they are wildly different! Stephen Fluin brought this to my attention. Take a look at a Miley Cyrus image search on Google and Bing.

Google Image Search: Miley Cyrus


miley cyrus google image search results

Bing Image Search: Miley Cyrus


miley cyrus bing image search results

As Stephen delightfully put it:

“Bing is like a loving grandmother who is in complete denial about her granddaughter’s behavior.”

Bing’s image search is completely devoid of Miley’s entire 2013 MTV VMA performance. The twerking, the tongue, the short haircut make no appearances. That performance is arguably the biggest splash Miley has ever made and Bing is pretending it didn’t happen?

Clearly the image search algorithms employed by Google and Bing are very different. Lets look at another pair of examples and see what we can figure out.

Google Image Search: Lindsay Lohan


Google Image Search Results for Lindsay Lohan

Bing Image Search: Lindsay Lohan


Bing Image Search Results for Lindsay Lohan

Google Image Search: Amanda Bynes


Google Image Search Results for Amanda Bynes

Bing Image Search: Amanda Bynes


Bing Image Search Results for Amanda Bynes

In all of the Bing results not a single “bad” or “negative” photo is shown on any of these three female celebrities. My from-the-hip guess is that Google values social signals higher. Images with a greater amount of social shares appear more frequently and higher than lower scoring photos. Mug shots, or explosive pop culture photos often get shared quickly and appear on many gossip news sites which score large social numbers. Bing might take a more traditional approach and show pictures from highly authoritative websites?

Let’s use the example of Lindsay Lohan. Bing’s top 10 images from from:

mantoos.com
fanpop.com (two times)
blog.screenweek.it
stud-center.com (four times)
fotoblogx.blogspot.com
iwritealot.com

I don’t personally recognize any of these. Following through some of the links, I wouldn’t consider these sites reputable. Compare this to the Google’s sources:

businessinsider.com (two times)
nydailynews.com
hdwallpapersdj.com (four times)
imdb.com
vogue.fr
perezhilton.com

IMDB, Business Insider, and Vogue? I think most people would agree that these sites are certainly more reputable and authoritative. So what is going on here? I enlisted the help of our Director of Inbound Marketing and search engine wizard Andy Forsberg to provide analysis. First he put together this spreadsheet comparing the top five Miley Cyrus images:

image-search

And provides the following analysis:

Bing images search places little to no emphasis on the date or authority of a given website when pulling images for its search results and sorts its images by their resolution (largest photos first). The higher resolution images tend to be professionally taken and are therefore, more likely to be more “appropriate” for a general audience.

Google image search places significant emphasis on the date and the authority of a given website when pulling images for its search results and sorts its images primarily by domain authority, page authority, social sharing metrics & recency. Since the more shocking photos are more likely to be linked to more frequently and shared more frequently on social media channels, they are more likely to show up near the top of Google’s image search results.

Clearly the algorithms the two engines use are very different. Of course, analyzing search engine algorithms is more than a full time job. The sample size we used here is not large enough to draw any real conclusions, only for fun speculation. So based on our limited research, which engine is better?

If you are looking for a high quality, beautiful photo of Miley Cyrus to hang over your fireplace, Bing will deliver. Using Google you would likely have to use advances search and set a higher minimum size and do some hunting before getting the perfect picture. If, instead, you keep hearing things about Miley Cyrus and you want to find out what the fuss is all about, Bing will be almost no help. The most recent image it provides is from May 2012. Google is better suited at bringing images from the latest events to the top. It is thought-provoking how the engines attempt to satisfy different needs when faced with a search like “Miley Cyrus” and no additional context. I am now curious how these tendencies manifest themselves when searching for things other than female pop stars…

If you have any other insights, or simply prefer one search to the other let us know!

bingiton1

Google vs Bing Analysis

Bing is performing, what I consider to be, a very brave advertising campaign called “Bing It On new window link icon.” It’s like the Pepsi challenge of our generation. The idea is simple: people use Google without ever feeling like they need to change. Bing challenges Google in a head-to-head search-off to see who truly yields better results.

bingiton thumb

I am going to come clean – I am a Google user. I have NEVER used Bing unless I am testing something SEO-wise. As a matter of fact, I am pretty sure all of MentorMate are pretty zealous Googlers. However, we are also technology agnostic. If Bing is a superior search engine we have no choice but to switch. I decided to run the Bing It On test to find out.

My testing process was systematic. First, I (with the help of my coworkers) came up with five common searches and discussed what our expectations of the results would be before running actual searches. We ran the searches to decide which results better met or even exceeded our expectations. Just for fun, we tried to identify which result was Google’s (something I will refer to as Google Signs). Once we began our testing process, we realized there were a few more common searches we were curious about so we extended the testing to ten total searches. As you’ll see below, each search is different and came with varying expectations.

Without further ado, bring on the test results!

Searching Something I Expect to Rank Highly

Search: “minneapolis software development”
SEO is one of our core competencies at MentorMate and we want to see the fruits of our labor. We work hard to get MentorMate.com and our client’s sites as high in the rankings as possible.

Expectation
I want to see five or more different mobile application development companies. I am curious which engine ranks MentorMate higher.

Results: Bing Wins
Both returned a nice list of software development companies. The results on the right, however, delivered more. The engine on the right also delivered a map and phone numbers. I like when my search engine goes above and beyond for me. You can tell that these are the Bing results because the map is clearly not a Google map. MentorMate is #1 for this term in Google and 4th or 6th, depending on how you look at it, in Bing. This makes sense because we spend less of our time optimizing for Bing, which makes me think… Are the companies above us optimizing for Bing or does Bing just value different things?

Searching to Spellcheck

Search: “mtstisfactoy”
This has probably become my most common usage of search. Looking through my search history will yield dozens of misspellings of words like necessary, entrepreneur, and diseases. In this case, I engineered a horrible butchering of the word ‘satisfactory’ by closing my eyes and attempting to type the word with the third knuckle of my left hand (I am right-handed).

Expectation
I expect to get a “did you mean satisfactory?” at the very top.

Bing vs Google analysis search term mtstisfactoy satisfactory misspelled

Result: Bing Wins
Google failed to find anything, while Bing gave me exactly what I needed. I ran this search again outside of the BingItOn challenge to see if it was some kind of glitch. Nope. Google really can’t figure out what I mean. This is very surprising because I search horrible misspellings all the time in Google and don’t recall stumping it outright. I ran a bunch of misspelled words after my original test and Google did fine in all of those. However, Bing did a perfect job 100% of the time while figuring out what I was trying to type so it gets the win.

Searching a Question

Search: “how to prepare for a zombie apocalypse”
Asking a question is classic; probably the most common reason to use a search engine. Recently, I remember asking “how to remove a stain from a tie”, “how does dry cleaning work”, and “how to best train my dog”.

Expectation
I want the top three posts to be informative blog, forum, or company links that sufficiently answer my questions.

Bing vs Google analysis search term how to prepare for a zombie apocalypse

Result: Bing Wins
Both engines returned worthy results from reputable sites. None were completely off-base, though they did range from funny to serious – but that is to be expected from such a question. I gave the win to the left side because it provided instructional videos while the right side did not. I was surprised that the left was Bing! I thought I was choosing Google because of the YouTube links. It’s pretty strange that Google decided not to display YouTube links for this search while Bing did.

Searching for a Movie

Search: “the adventures of baron munchausen”
Sometimes you want to learn who stared in a movie, what the DVD cover looks like, or see a trailer.

Expectation
I want to see a trailer, a link to the IMDB movie page, and a link to the Wiki page.

Bing vs Google analysis search term the adventures of baron munchausen

Results: Google Wins
Both engines delivered exactly what I needed. Google is on the right, obvious from the YouTube thumbnail in result three. Notice how Google displays star ratings from IMDB and RottenTomatoes? One of the core pieces of information from those sites is delivered without needing to waste a click. Bing tries to bring something special by displaying the site thumbnail images. Unfortunately, that does not stand up to Google’s increased usability and Google gets the win.

There is more to this story. Google’s Knowledge Graph. The BingItOn tester excludes any superfluous sidebar information. That sidebar is where Knowledge Graph information appears. Take a look at this same search on Google. It is a complete blow out! Look at all the information I get about this movie without clicking any links. Even if Google lost in the BingItOn test, it still would have won if the Knowledge Graph was taken into account.

Searching for a Movie by Description

Search Term: “low budget time travel movie”
Let’s say you forgot the name of a movie or video game, but know what is about.

Expectation
I expect to know the name of my movie, and get an IMDB and Wiki page as well as a trailer link.

Bing vs Google analysis search term low budget time travel movie

Result: Google Wins
They both correctly identified the movie. However, only Google returned the IMDB link. I find it strange that Bing decided not to pull IMDB even though it did in the previous search despite identifying the movie correctly. Google gets the win for this one because of the star ratings and, of course, the IMDB link.

Searching for Sports Facts

Search: “most homeruns this season MLB”
Sometimes you just need to settle an argument.

Expectation
I want to know the answer to my question without having to click on another link. I want the link title or intro text to answer it so that I can go on with my day.

Bing vs Google analysis search term most homeruns this season mlb

Result: Draw
Both search engines fail to answer my question without forcing me to click links. Digging deeper, I still worked entirely too hard to get the answer to my question. Maybe this is due to me being an amateur at navigating sport websites? Interestingly enough, Apple’s Siri was able to tell me exactly what I needed: “44 homeruns by Miguel Cabrera.” So I guess Apple wins.

Searching to Convert Units

Search: “how many cups in a liter”
This could be distance, volume, temperature, or even money.

Expectation
I want the results to my conversion without needing to click on any links.

Bing vs Google analysis search term how many cups in a liter

Result: Google Wins
Both returned the correct result in the way I expected. Turns out Google, on the right, brings extra usability to the table. It offers an interface where I can change either number and see real-time conversions without needing to search again. This feature must have been added recently because I do not remember ever seeing it. But for the increased usability, Google gets the win.

Searching a Business

Search: “Umbria Pizzeria Bloomington”
Sometimes you need to call a place to make reservations. Or write down the address. This process should be very quick and easy.

Expectation
I need to get an address and a phone number in the top three results, preferably without being forced to click any links. Then I would like the business website and some reviews.

Bing vs Google analysis search term umbria pizzeria bloomington

Result: Google Wins
Obviously, the left side wins with flying colors here. We can tell it is Google by the familiar review stars being displayed in the results. Another Google sign is the internal links it shows for the top result. Of course, the left result doesn’t win just because it is Google; it wins because the address and phone number are right at the top along with a “contact/map” link. Also, Bing doesn’t return any review sites while Google not only brings a Yelp review, but displays the star-rating without any extra effort. Take a look at Knowledge Graph information for an even more decisive victory.

Searching for News

Search: “japan riots in china”
Did you overhear something on the radio just as you arrived at your destination that you want to learn more about? Maybe you are fact checking some current event to make sure you aren’t only getting one side of the story?

Expectation
I would like to get relevant news posts from a variety of sources.

Bing vs Google analysis search term japan riots china

Results: Draw
Google brings us Examiner, News Asia, AFP, and the Globe and Mail as sources. Bing brings Bloomberg, CNN, Rueters, Yahoo Finance, and HuffingtonPost. While both returned a variety of sources and images, Google goes the extra mile and embeds videos. But both results pages met expectations and a YouTube video isn’t immediately a relevant news source. I give this search a draw.

Searching to Shop

Search: “mens dress shirts slim fit”
When I need to buy an item, searching is often my first stop. It helps me price check, informs me of which stores carry the item, and even provides me with some style suggestions.

Expectation
I expect to see images of the item so I know the search is understanding me correctly. I would also like to see some prices listed in the search page so I don’t need to click on any links to get a general idea of pricing. I would also like multiple stores which sell my item.

Bing vs Google analysis mens slim fit dress shirts

Results: Bing Wins
This was very surprising, but the left results are actually Bing’s! That’s right, Google returned 0 images and hardly any prices in the results. Google did a great job bringing a variety of retailers which sell slim-fit men’s dress shirts, but the results page is devoid of life. Bing meets my expectation by showing me images of shirts with attached prices. Although the variety of Bing’s results are more limited, I still prefer the result including images and prices.

But wait! This is another example in which Google’s hidden Knowledge Graph devastates Bing. Google does show images and prices, it just places them in the excluded Knowledge Graph. Comparing the images, it looks like Google does a better job actually showing “slim fit” shirts instead of just dress shirts. If I could factor this into the test Google would win this search, but I am limited to an honest BingItOn test.

Bing vs Google analysis final results draw 4 to 4

Conclusion

I am very surprised to see this end in a draw. But, it is clear that comparing search results alone is not comparing apples to apples. Two of the search results are suspect. When I searched for a typo, I somehow won the lottery in finding a word Google couldn’t figure out. 99.9% of the time that result will be a tie. Also, the men’s dress shirt search in Google actually does deliver superior results via their Knowledge Graph, which the BingItOn test appears to block. In my opinion, displaying star ratings in the results, interactive widgets like unit converter and calculator, and the useful and impressive Knowledge Graph empower Google with usability that goes beyond simply retrieving relevant links. Taking that into account, Google has the edge. While Bing is a fine tool, I am in no hurry to switch primary search engines just yet.

links1

Link Building with the Experts – Our Link Building Tips & Strategies

Outspoken Media just interviewed 11 famous SEO experts to answer 11 great questions about link building. We saw that Dori had gone through and answered these link building questions on her blog, so we thought we’d do the same. We hope you find these answers useful to your own link building campaigns.

1) What are a few emerging link tactics that you’ve seen in the past 12 months providing tremendous value to sites/pages? Can you give a specific example or two?

I haven’t found any new link tactics in the past 12 months that provide tremendous value. All the great tactics that truly provide tremendous value have been around for a while.

2) The SEO industry has become so stingy with linking to quality content to the point that many people who used to share a lot of it simply don’t bother, as it is not worth the cost of doing so. Is this a trend which spreads? Are we canaries in the coal mine, or is this just an issue impacting the SEO niche because it is far too saturated? What can Google do to encourage organic linking on the WWW (outside of nepotism, hype, spin, misinformation & ego-baiting)?

We don’t like to look greedy, so we don’t perform the stingy acts on our company web site. However the reason we created SpyderMate initially was as a linkbait tool. So for SpyderMate we don’t freely post external links. We feel it’s justified there however because we are offering a free and very useful service to the community.

3) What are the criteria for the “perfect link”?

The perfect link has the following attributes:

a) High authority domain & page
b) Minimal amount of external links on same page
c) Web site is very relevant to yours
d) The anchor text matches a keyword you are battling for
e) It’s located within the obvious content section of the web page
f) obviously it’s not tagged id=”linkage”

4) How do you go about creating a link marketing plan that will A:) Get tangible search results in a 6 to 12 month period and B:) Create sustainability for the website you are creating the plan for (i.e. keeping the links clean and adding links with long term value)

Well first you find out what keywords they should be targeting. Then you go through all their competitors ranking  for the keyword phrases and scour all their backlinks. Once you make sure you got a link from every backlink they have possible then you move on. Next you target the more lower hanging fruit longtail keywords through your list of standard directories that you submit to. This helps re-enforce your short-tail keywords with long tail variation power. After this you go for the seriously powerful links. These are your premium directory listings, your serious linkbait web sites and your powerful friend’s web sites. These sites target the more generic short-tail keywords that help raise all your longtails associated around them. The maintain sustainability you consistently create great new content and linkbait to create a consistent flow of natural inbound, quality, relevant links.

5) If you could choose a link on a lower authority page that would provide a moderate amount of targeted traffic or a link on a higher authority page that would provide absolutely no traffic – all other attributes being equal – for ranking benefits on the site you’re developing links for, which would you choose and why?

This depends entirely on how relevant and how well converting the lower authority page would be. It also depends on current rankings, whether or not there is a keyword I could boost my rankings for that would beat the conversions coming directly from this link. So if i had a site with a low conversion rate that had a high payout I would go with the lower authority link. If I had a site with a high conversion rate and a lower average payout I’d go with the high authority link.

6) Do you feel that you can conserve pagerank or that it’s still worth the effort to sculpt your links, by limiting the number of links on a page, creating them with JavaScript, passing them through a blocked page or using nofollow?

I believe you can, but also don’t believe it’s worth the effort, especially on a site that you’re constantly updating. There are other tasks you can spend your valuable time on that provide more value to your web presence.

7) Please discuss what link deprecation is and what impact it may have on a link building campaign.

I believe link deprecation to be the result of poorly built links primarily. The only reason this should occur is due to the age of the links. Actually losing links only happens to links that were easily attained and therefore easily lost. The best links are very hard to come by and last the test of time.

8) Do you think search engines are trying to find a way to depend less on link popularity and more on other algorithmic/social media factors?

I believe they are, but I don’t think it plays much of a role currently. I also think basing searches on social media is not viable for many categories of web sites. The fact that social media sites like Facebook are social make your likes inherently based around your friend’s general interests as well as your own. Even though I’m heavily interested in SEO, it does not mean I’m going to bother liking SEO articles I find interesting left and right because I know it will annoy my friends. I want to share things that I know my friends will also be interested in. This is not a system that a true search engine can be built upon. A search engine can however leverage this information is done on the right level.

9) How much do you stress internal linking on your own or clients’ sites? Do you have a quick rule of thumb or strategy to maximize the effectiveness of internal links?

I’ve found that emphasizing your traffic generating pages through inbound links is critical. Using carefully placed inbound links to associate relevant topics and re-emphasize targeted anchor text is extremely helpful, especially for that highest level page in the link scheme. Enough said.

10) What’s a successful link development strategy many overlook or dismiss?

Linkbait. People tend to dismiss serious linkbait strategies because of the time/costs associated with them. They tend to think other efforts are easier, but nothing generates those truly unique, critical, powerful links like well done linkbait.

11)  What have you been most WRONG about over the course of your link building/SEO career?

I was most wrong about how effective thousands of links coming from the same authoritative web site, with different anchor text, coming from different pages, linking to different pages were at lifting my rankings. Domain diversity is absolutely crucial. I’ve learned that one great warrior is better than thousands of mediocre warriors.

seo news1

Latest SEO Industry News

9/21/09 Google officially announces that meta keywords don’t impact ranking in Google

Matt Cutts just officially stated that Google does not use meta keywords to impact web sites’ rankings in Google. It has been widely known for a long time that meta keywords did next to nothing, if anything in Google. Cutts decided to make it official. I have always recommended to not use meta keyword tags on web sites because they are a waste of time and competitors can then see what keywords you are targeting. Purposely using meta keywords to trick your competitors into thinking you are targeting certain keywords is still an option I guess, but that seems like a waste of time as well. Any good competitor will be able to figure out the truth anyway, especially with tools like compete.com around.

9/17/09 Rand Fishkin of SEOMoz changes stance on paid links

After a lengthy period of pushing paid link services all over the SEOMoz web site, Rand Fishkin has decided to change his stance on paid links. That is, of course, paid links that are not properly deemed as advertisements and actually pass along PageRank. John Chow’s blog was targeted by Google for this practice awhile ago. As a result, he did not rank for his own name in Google for a long period of time. Google received a lot a negative response for this move from well-known bloggers and recently removed the punishment, allowing Chow to rank for his own name again.

Aaron Wall of SEOBook had some interesting thoughts on this move by Fishkin. Wall recently stated on his blog “If you philosophically didn’t believe in buying links then why would you spend $1,000,000+ building a web graph of link data? What good is researching all the link data if you take link buying off the table as one of the options? Most of the competing links that you can replicate will require some level of payment.”

7/29/09 Yahoo gives up on search, makes deal with Bing

While most of you already know about this, it is a major news item for the SEO industry in particular. This means no longer trying to rank in Google, Yahoo! and Bing. Now there are only two major search engine you have to focus your SEO efforts on. I do not know how to feel about this at this time. I am curious however, to see what kind of market share Bing has in a year or so.

QuadsZilla at SEO Black Hat seems to be very excited about Bing. He recently wrote a post stating “We want Bing to succeed: we need a real challenger to Google’s monopoly” and even goes as far as to get a contact at Bing in order to “negotiate a deal to bring upwards of 10,000,000 new searchers per year to Bing.com.”

bad website1

19 Signs Your Web Site Was Poorly Designed and/or Developed

1. You can’t find your own site on Google

If you’re looking for your own site and still can’t find it, something big is wrong. This is worst case scenario for a web site; if you can’t find your own site, none of your target audience will be able to either.  If you’ve typed in your company name or something else specific to your brand you should be at or near the top of the results.

2. Your visitors aren’t doing what you want them to

Your site has a specific purpose, whether it’s sales or to generate a lead through a contact form. If your site is receiving traffic, but that traffic isn’t using the site like you intended, there are probably issues with the way your website targets your audience.  In other words, you need to refocus your SEO to allow the right people to find you. Another problem may be with your web site’s usability. If visitors are unable to use it as you intended, your site has some serious problems.

3. Browser incompatibility

This is a design flaw that isn’t always noticed, but is crucial for the user experience. Your web site may look great in the browser you personally use, but different browsers render sites differently. Your site may look great in Firefox, but terrible in Safari.

4. Duplicate or nonexistent page titles

Page titles, the text shown in the top left corner of your browser, are one of the biggest factors in how a search engine finds your site. Each page’s title should accurately reflect the keywords you’re targeting for that specific page. Your pages won’t appear anywhere near the top of the search results without effective page title usage.

5. Excessive use of images in place of text

Text rendered as an image can often look more appealing to the eye, but it can hurt your search rankings.  If you use images in place of text, you’re missing a major opportunity to help search engines figure out what your site is about.

6. Main navigation isn’t persistent

Certain page elements should be present on every page.  The most important is your main navigation.  This might be tabs that are at the top of every page, or a list of website sections on the side of the page.  Regardless of how users navigate through your website, you should never leave them guessing about how to get to more of your content.

7. Excessive use of Flash

Creating a site exclusively with Flash can look impressive, but it is nearly impossible to rank highly in search results by doing this. Flash can be used to enhance a site, but it shouldn’t make up the entire site.

8. Websitegrader.com gives site a score less than 25

Website Grader is a free tool that measures your site’s marketing effectiveness. It analyzes the amount of traffic you receive, your SEO efforts, social networking popularity, and technical factors to produce a score from 1-100. Scoring below 25 would signify major problems in several areas of your site.

9. Intro pages

An extensive introduction may look nice, but intro pages can also detract from the user experience.  This is especially the case for your most valuable user – the repeat visitor – who may not want to sit through the same intro every time they visit your site.

10. Page URLs aren’t human readable

Your URLs should be simple and say what each page is about. For example, www.dogs.com/dalmatians is better than www.dogs.com/show.php?breed=11. This helps search engines better index your content and your users stand a better chance of remembering how to get to your content.

11. Site contains excessive text

Too little text can keep search engines from understanding what your site is about, but too much text can drive away visitors. Follow the rule of thumb that nobody on the web wants to read; users should be able to quickly find and use the information they are looking for without reading anything unnecessary.  Limiting your text also helps you focus on the most important words, too, and this helps search engines figure out what your site is about.

12. No contact information is present on the site

This is obviously easy to remedy, but also a problem that some sites forget about. You need contact info to receive both leads and feedback on areas to improve your site.

13. Slow load time

Internet users are impatient. The longer they wait, the less likely they are to use your site. Significantly slow load times can also negatively impact your rankings in the search engines, resulting in less web traffic to your site.

14. Your site is designed for small monitors

As computer hardware advances, your website needs to also.  A site designed for 800×600 screen resolution may have been state of the art in 1997, but today users expect your site to utilize their entire screen. A width of 960 pixels is optimal these days.

15. Canonical URL issues

http://www.sampledomain.com and http://sampledomain.com (no www) should 301-redirect to one variation of the domain name. You need to choose with or without the www. This way you can ensure no link juice is wasted on the less popular variation of your domain.

16. Too many navigation choices or too many clicks to get there

Your site has to be easy to navigate. Too many options on the home page are overwhelming; will spend too much time reading through all the options.  At the other end of the spectrum is excessive page depth.  Forcing users through a seemingly endless train of pages is a sure way to get them to leave.  As a rule of thumb, users should never have pick between more than 7 navigational options per page and they should be able to get to any page in at most three clicks. All pages should also link back to the home page.

17. Single column display format

This is simply an outdated feel and looks unprofessional as well. Make use of all the space presented to you. Single column formats lead to excessive scrolling. Don’t make extra work for the user.

18. Site has no Compete.com data (after six months of existence)

Compete.com is an online tool used to measure traffic against your competition. If your site has existed for a substantial amount of time and has yet to be noticed by Compete, your site is likely getting very minimal amounts of visitors.

19. There are specific on-page browser recommendations

These should be unnecessary because as previously stated, your site should be usable in all browsers.