Do Search Results Vary By Browser?

February 9th, 2010 • By:  • Search Engine Optimization

browsers 1About a week ago I got involved in a discussion over on LinkedIn at the Search Engine Land group that was started by someone asking the question:  “Google showing different search results for IE and Firefox. Any explanation?” I gave my opinion which was that most likely it was personalized search impacting the rankings because one of the browsers was their “primary” and would have all the cookies and history, while the other would not. But some other possible contributing factors were also presented like “were you logged in to Google or not” or “was Google toolbar installed” and a few others.

So, after following the discussion a while I decided to put the topic to the test and actually run a series of tests (10 all together) to see if in fact Google rankings are different depending on the browser (or a combination of other contributing factors). Interested in the results? Read on to find out!

The browsers and versions included in the test are:

Firefox: v3.5.7, Internet Explorer: v8.0.6001.18828, Google Chrome: v4.0.249.78 (36714) Safari: 4.0.4 (531.21.10) 

Keywords:

I needed a couple unsuspecting keywords… I figured it best to use more than 1 keyword just to make sure the results were not impacted somehow by the keyword in question.  I chose “Indian motorcycles for sale” (the bike I learned how to ride on was an Indian) and “android phones” because I LOVE my new Droid Eris.  Worth noting: I have not recently done any Google searches for either of these keywords.

The 10 Tests:

Below are the tests I performed.  For clarification purposes “as is” means my browsers in the current state they were in. Nothing changed, cleared, modified. Just open and go.  And, for the record I use Firefox 98% of the time so it had tons of history, a number of toolbars installed and lots of cookies. Also, you will see references to the “personalization query” in the tests. In this instance I appended “&pws=0” to the end of the url Google generated, manually telling Google to turn off personalization for this query.   As you may or may not know, personalization is "on" by default now. That’s why it was important for me to use this command and ensure true results.

  • Test 1:  Google each keyword using google.com on all 4 browsers “as is” signed out of Google with Google Toolbar installed, record the top 3 results
  • Test 2:  Google each keyword using google.com on all 4 browsers “as is” signed in to Google with Google Toolbar installed, record the top 3 results
  • Test 3:  Google each keyword using google tool bar on all 4 browsers “as is” signed out of Google, record the top 3 results
  • Test 4:  Google each keyword using google tool bar on all 4 browsers “as is” signed in to Google, record the top 3 results
  • Test 5:  Google each keyword using google.com on all 4 browsers “as is” signed out of Google with Google Toolbar installed, using the personalization query
  • Test 6:  Google each keyword using google.com on all 4 browsers “as is” signed in to Google with Google Toolbar installed, using the personalization query
  • Test 7:  Google each keyword using google.com on all 4 browsers with the cookies and cache cleared signed out of Google with Google Toolbar installed
  • Test 8:  Google each keyword using google.com on all 4 browsers with the cookies and cache cleared signed in to Google with Google Toolbar installed
  • Test 9:  Google each keyword using google.com on all 4 browsers with the cookies and cache cleared signed out of Google with Google Toolbar uninstalled
  • Test 10:  Google each keyword using google.com on all 4 browsers with the cookies and cache cleared signed in to Google with Google Toolbar uninstalled

The results:

Can we have a drum roll pleaseeeeeeee………I got exactly the same results every time with every browser regardless of the conditions!  Disappointing huh?  I was really hoping to have something to talk about after the tests but I don’t.  That’s it!

Conclusion:

For now,  all browsers seem to be serving up the same results in Google so rest easy at night knowing that although you may have to test your website design with different browsers you don’t have to test your SEO with different browsers.

Have you found differences in your search results between browsers? What tests have you run?

Tags: , , , , ,

This entry was posted on Tuesday, February 9th, 2010 at 6:25 am and is filed under Search Engine Optimization. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

11 Responses to “Do Search Results Vary By Browser?”

  1. Kyle Kazak Says:

    Chris,

    Good experiment. I have heard that people have been presented with varying SERPs depending on Web browser but like you said it always seemed to be cookies or personal settings that created the variation.

    There is one way that Google will display different SERPs but it does not involve Web browsers, rather different data-centers. I have built a Google keyword tool that is hosted in Dallas, TX. The data it pulls back is from the nearest data center, which happens to be in Houston, TX in this case. Our office in Northern California yields similar but different sets of data. We have also tested this theory from Colorado and New York. I am not 100% sure if it is lag time between the different data centers or what but they definitely yield slightly different sets of data.

    We originally used Google’s search API to perform the query but decided that it has numerous known problems. Our scrape/IP method is far more accurate but without knowing every Google Data Center location/IP we are unable to route the user’s query to their local data center. Hopefully one day Google will fix this!

    Let me know what you think.

  2. Chris von Nieda Says:

    Hey Kyle,

    Thanks for the great comment. Seems you have taken it a step further then what I did, great work! I’d be curious to know when you say different SERPs just how different? Are we talking one or 2 positions or huge swings?

  3. Kyle Kazak Says:

    Chris,

    In most cases where the website has been online, verified, and sitemap submitted, the positions only differ no more that 1-2 places.

    If the sites have been ‘sand-boxed’ or whatever Google claims the sifting period for new sites is, I have had the data center in Houston, Texas ping back position 1-10 and when I actually do the search it is un-indexed, no where to be found. A few days later back in the 1-10 spot.

    Huge swings only tend to happen with new sites, especially ones with less authority.

  4. Chris von Nieda Says:

    Gotcha, well I guess the mystery known as Google lives on. First one to solve it wins.

  5. Double Vee Says:

    Hi there.
    Thanks for that experiment and for sharing the results. I was wondering if you had also done some testing on mobile search, using your much-loved Android?

  6. Chris von Nieda Says:

    Hi Double Vee,

    Thanks for your comment. No but that would make an interesting follow up experiment. Are you suggesting compare results of different mobile browsers installed on the same device or the different default browsers that come on the various devices like iPhone compared to Droid?

    - Chris

  7. Double Vee Says:

    Hi Chris,

    I was more thinking of the first instance: comparing different mobile browsers set up on the same device.

    But it might also be interesting to cross-check results obtained from mobile browsers (with the geolocalization feature turned off on the mobile device of course) to standard SERPs (from your regular computer)…

    Best regards,

    VV

  8. SEPo Says:

    Hi Chris

    One thing I have noticed over the past month or so is clients in different parts of the Uk are getting different results to me for the same keyword – sometimes as much as 5 places difference.
    Like you – at first i thought it was down to personal search but they have all disabled their PS.

    I’m assuming this has a lot to do with the proposed Google Caffeine update which no one from google has commented on since they said it would go live after xmas hols.

    Cheers Jim

  9. Chase Says:

    Our site, being slightly over a year old, is dramatically effected by what appears to be data center variances.
    On one browser from a computer I use commonly, we show high in several keywords. On the same computer with a different browser, used less commonly, we don’t show at all.

    The variances from a mobile device and from less commonly used computers seem to side with the lesser used browser where we appear infrequently if at all. It’s strange behavior and for my sake, I hope it normalizes :).

  10. James - Traffic Travis Says:

    Good test!.. Havent' heard this before where different browsers vary G search results. By theory, you could say its probably due to cookies enabled. Great to see the test results disapproving.

    Well, another addition to this test would be, querying a number dummy keywords first with related niche of the test keyword having cookies enabled and disabled on different set of browsers. Then try to click some consistent theme search results, i.e, query words like apple, then click in search results considering apple as a fruit, not a gadget.

    This results may or may not contribute to some varying search search results with different browsers.

    James

  11. Dave Says:

    My website shows different results between firefox and IE for the same keywords ” computer repair
    London” and the results also vary between computers, I have not even done a test on other browers.
    My concern is that even if the web experts know how to tweek the browers to get the same results this will have no effect on our target audience.

    It is a real mess for website designers and promotors. Personal settings should never inlfuence search results. this is the root cause but it is still a level playing field.

Leave a Reply