.Com Versus New gTLDs: Real World Test Results

Are .Com domain names truly better than the new gTLDs? Well, I have had my suspicions, but there was no way I was going to base my opinions. I decided to put some newly acquired domain names to the test, and use Google AdWords to reveal the results. While there’s no really easy way to test the organic search results, using Google AdWords to drive traffic and test conversions was much faster and more reliable.

I set up two tests. One test used a .Com and a .Diamonds domain name. The other used a .Com and a .Menu domain name. In the first test, I was able to secure two keyword rich domain names: one with the keyword in the domain name, and the other with the keyword in the domain name and in the new gTLD.

I chose these domain names for the primary test:


and I chose brand-related keywords for the second test. I chose these domain names for the second test:


I set up two separate landing pages. One landing page for the “diamonds” test and another landing page for the “menu” test. In this case, it was important to make sure that everything was the same for each test. So, everything was exactly the same, including the ad copy used, the landing page, and even the conversions. The only change between each test was the domain name being used.

For each Google AdWords campaign, it was important to us to not only use the same ad copy in the ads, but also to make sure that we used the same keywords in each of the campaigns. We did some preliminary keyword research, and decided on a list of keywords that had enough searches per month to give us enough data to work with.

For the diamonds Google AdWords campaigns, we used 50 keywords total, all priced at $1.00 per click (that is what we bid, but not what we ended up paying for each click). We also bid $1.00 a click on the .Com versus .Menu domain name test.

The Results

For the .Diamonds versus .Com domain test, the average cost per click was $.81 cents per click on the .Com keywords. However, it cost us on $.77 cents per click when we used the .Diamonds domain name.

Average Cost Per Click: CPC across all keywords:
3Carat.Diamonds: $.77
3CaratDiamonds.com: $.81

It ultimately cost less per click for a .Diamonds domain than to run the same keywords on a .Com domain name, and the total campaign cost was lower. With a higher CTR on the .Com domain name, it appears that end users favored the .Com domain name over the .Diamonds domain name.

The .Diamonds domain name, however, was given quite a few more impressions than the .Com domain name, giving the .Diamonds domain name more visibility. In fact, it appears that Google AdWords actually favors use of the .Diamonds domain name, giving it more impressions and even better positioning. The average position for the .Diamonds domain name was better. If we look specifically at CTR, the .Com domain name appears to be the winner here.

When it came to conversions, we saw a higher overall conversion rate for the .Com domain. It was nearly 20 percent higher than the .Diamonds domain name. But those conversions came at a higher cost, as the cost of the .Com campaign was higher.

Here are some other key takeaways from our testing:

  • The effective CPM of a .Diamonds domain name is essentially half of what it was for the equivalent .Com domain name. So, if you’re in the jewelry industry (specifically diamonds), there may be a great opportunity to get twice the impressions for your money.
  • In the .Diamonds versus .Com test, the Average CPC was $.77 cents and $.81 cents per click. However, we got more conversions using the .Com domain name.
  • It appears that Google AdWords tends to favor the New gTLDs, especially because they gave the New gTLDs more impressions for a cheaper cost, and gave ads better positioning overall (.Diamonds vs. .Com).

So, at this point, consumers prefer the .Com. Google AdWords, on the other hand, prefers the .Diamonds domain name.

We’ve published all of the results, including all of the ad copy, landing pages, conversions, and even the keyword data in our report, available on Globe Runner’s website.


  1. I have serious doubt regarding the conclusion from this result. Here are few important questions needs to be answered before making any conclusion.

    1) How long this test was conducted?

    2) What was the total budget spent on the test?

    3) Was this test done under controlled environment? Like Running ads for same time duration, same targeting etc

  2. Dheeru,
    The test was conducted over about a month’s time. The total budget for all of the campaigns are listed in the final results document/white paper that you can download on the GlobeRunner.com website. Each campaign had a specific budget.

    I’ve also included all of the details, even the keywords, CTR, CPC, bid prices, average positions, everything in the document. So, to answer your number two and 3 questions above, it’s all in the PDF file.

    • Hi Bill,
      Thanks for your reply. I really appreciate the work you guys have done.

      I went through the document and have a few questions for you (I hope you don’t mind πŸ˜‰ )

      a) Do you really thing that a budget of $68 for each domain ( .diamond and .com) is sufficient to make a conclusion?

      2) Do you think a difference in CPC of 4 cents make a domain extension winner or loser?

      3) Don’t you think that such kind of experiment should be done in a more controlled manner and should be done to check the overall search engine impact on these new Gtlds (Testing Adword and SEO)

      My only concern is regarding the impact of these kind of results- I am a part time domain investor and have seen newcomers trying to make quick money and they end up in loosing.

      With these kind of experiment and results, new GTLds will become more favorable from Domain investing point of view and ultimately make these registries rich and their investor poor.

    • Dheeru,
      I’ve tried to answer your questions below each question:

      a) Do you really thing that a budget of $68 for each domain ( .diamond and .com) is sufficient to make a conclusion?
      We were going for 100 clicks per campaign, so it only ended up at $68 over about one month’s time. We originally planned on $1.00 per click with a $100 budget. It just ended up that way. We’re continuing with our testing so we have more data–this is just the amount of data that we decided to release publicly. I agree with you, though, the more data that you have the more informed decision that you can make (with any test).

      2) Do you think a difference in CPC of 4 cents make a domain extension winner or loser?
      It really was very close. What’s important to note here is that it WAS SO CLOSE. That shows that a new gTLD is “just as good” as a .Com domain name. If it were NOT close then we would be seeing that either a .Com is better or worse than using a new gTLD domain name. Again, what’s important to note is that the CPD was very close.

      3) Don’t you think that such kind of experiment should be done in a more controlled manner and should be done to check the overall search engine impact on these new Gtlds (Testing Adword and SEO)
      Since there is no controlled way of testing SEO (there are so many SEO ranking factors and you cannot simply have all the same links and all the same content on the same site), we had to do a Google AdWords test. I would like to do an SEO test, and I’m working on that. But again, you cannot control everything down to the domain name only if you’re doing an SEO test. It has to be a PPC test.

  3. Thanks, Bill. Nice to see your work. I’m particularly curious to see how these results continue to skew when Google and Amazon launch their TLDs in earnest later this year. Just wait for the 1st Fortune 100 .BRAND to launch and run a real campaign changing consumer behavior. We are still in “pre season” scrimmage mode, so the real games have yet to begin. Interesting times.

  4. Eric,
    If you take a look at the PDF file I outline everything, including what the conversions are. For the “diamonds” test, it was downloading a PDF or clicking on a “shop now” type of button to go on to another ecommerce site that sells diamonds rings, etc.

    For the “menu” test, we were looking primarily at time on site and whether or not they viewed a menu or not.

    Again, if you look at the test results, everything is outlined in the white paper.

  5. It sounds like a great experiment, but the one problem I have with it is how you define a conversion. Although the way you did it would be fine for a lead generation site or landing page, the vast majority of sites who would be interested in these results would actually be selling something, and the domain name the site uses might make much more of a difference. For example, the actual diamond sales are at sylviecollection.com. If that was sylvie.diamonds instead, would it get as many sales? In other words, would people trust a .diamonds enough to actually hand over money to them? I am not saying they wouldn’t, but that seems to be a much more important thing to test.

    • Eric,
      We chose the domain names that we did because we didn’t want any “brand influence” to play a role here. We wanted consumers to click and convert based on the TLD, not necessarily anything else that influenced them.

      For those who clicked the “shop diamond rings” button on the landing page, they were brought to the ecommerce site that. We aren’t reporting publicly how many of those conversions ended up in actual sales.

      Another aspect to test, though, would be to send the clicks directly to an ecommerce site that sells the products… and that may be our next test. In that case, though, we would need to set up two separate ecommerce sites on two different domains: one on the .Com and another on the gTLD.

  6. The most important thing was the much higher CTR that .com had and secondary the much higher conversions it had.
    This to me shows that .com is more recognized and trusted.

    The difference in CPC is non existent.

    Impressions may or may not be important depending on the campaign. If you want to actually sell something then impressions don’t matter. CTR does.

    Maybe google was trying to spend your budget on .diamonds and because noone was clicking on it, it was serving more and more ads.

  7. My takeaway from it was that the 2 TLDs actually did much much better than I thought they would. But, that was in a lead generation type situation, which is not something I am involved in, so I am unsure how that applies to actual sales.

  8. “We were going for 100 clicks per campaign”

    Completely useless test that would get laughed out of any affiliate marketing forum or anybody that actually did this for a living. Simply not enough clicks to determine anything. You should have waited to release anything until that happened.

  9. Jonathan,
    The data is what it is. This is the amount of data that we were comfortable with releasing publicly. We have more data from our tests that we’re not releasing to the public.

    As with any test, one could always argue that there’s not enough data. It’s a moving target… you start and adjust as you go along.

  10. Investment value for domain name extensions is clear. If you look at the value, as of today, between .com and .net, you can see, The last auction Sedo conducted for numerical domain names in comparing the 4 numbers .com domain names and the 4 numbers .net domain names, you will see while the least .com domain name was sold for $7,242, the .net domain name was sold for $230. This puts .com 20 times more valuable than .net. If you are going to buy the new top level domain names, don’t expect a fast profit, no matter how good the domain is. If you are the end user, and you are going to use the domain for a website, the new extensions are as good as .cc, .ws or .bz and so on. You can’t sell them for profit but you can definitely use them for websites, and don’t expect the same respect from your customers that you would get if you had a .com extension.

    1029.com 5,600 Euros ($7,242)
    1956.net $230
    1151.com 13,100 Euros ($16,697)
    5198.com $20,000

  11. I have 2 dogs, a pure bred and a mutt, I have had my suspicions whether the pure bred was a “better” than the mutt, so I did a test to measure the amount of hair each dog shed and I determined the mutt sheds the least, he is clearly the “better” dog.

  12. Before everyone goes out and starts a spending money and promoting this, I would suggest broader testing or analysis needs to performed, or a larger audience be surveyed on their experiences.

    Google is constantly changing, while it may work today, it may not in the future.


Please enter your comment!
Please enter your name here

Recent Posts

Rookie Mistake: Reading Expiry Lists at Breakfast

Every morning, as early as I am awake, I look through domain name expiry lists to see what is coming up for auction. I'll...

GoDaddy Auctions Running NameFind Auctions

GoDaddy is running a featured auction via its GoDaddy Auctions platform with domain names from its NameFind portfolio. While I would argue the domain...

Spaceship Hits 1 Million DUM – Only 13% of New Registrations are .com

Earlier this morning, Richard Kirkendall shared that Spaceship hit the 1 million Domains Under Management (DUM) mark. Richard is the Founder of Namecheap and...

Converse.CO UDRP Decision Turns on Price Inference

In general, I thought UDRP panels have gotten past the issue of pricing as it relates to generic / descriptive one word domain names....

Ebbs and Flows of Domain Investing

My domain investment portfolio has grown from around 500 domain names to around 2,000 domain names in the last several years. Even at that...