Myth Busters: Does a Domain Name Matter? (28)
STM MythBusters - Episode #1: Does a domain name matter?
Lately I’ve seen a lot of the question “Do I need a separate domain for each campaign/vertical/my company?”
I always answer “yes, probably”. Buuuut this is just an assumption.
How about we let the stats do the talking instead…
The Test
Send $100 of redirect mobile traffic to the exact same simple lander on 5 different domains.
The competitors: from my assumptions of best to worst
The offers
Two low payout US utility offers; DU Speed Booster & Apus Launcher.
My assumptions
Obviously there’s no way anyone is going to click through a lander with “scam” in the domain name.
There will be a huge difference between the related domain names and the scam names – with the random domain (4e7g383.com) sitting somewhere in the middle.
I’m pretty sure most of you would be with me on this one.
Results
What’s the statistical significance of these results you might ask?
We can be 77-86% confident that both of the related domain names (android-tools-today.com and top-apps-store.com) will outperform both of the “scam” domain names (free-tool-scam.com and scam-website.com).
We can also be 73-76% confident that the irrelevant domain name (4e7g383.com) will outperform both of the scam names.
Conclusions
People do take notice of your domain name, and this CAN affect on your bottom line.
BUT although there’s a difference, in my test it didn’t seem to be huge one.
This was the weirdest thing for me. I thought the word scam would keep anyone away - but perhaps not that many users bother to look up at the domain name as they’re too focused on the content of the site itself.
Looking past the numbers, there are also other reasons to diversify your domain names (including company name) – as have been discussed elsewhere on the forum…
Pros for having multiple domains
- Harder for affiliates to rip your other landing pages (your landers won’t all be sitting there like a LP goldmine waiting to be raided by other affs…)
- Your company isn’t tied to specific offers in the case of dissatisfied customers
- Company privacy – maybe you don’t want others knowing about your adult dating ventures etc.
- Better continuity with the offer through customized domains
Cons for having multiple domains
- Harder for advertisers to find your company and contact you directly
- (Small) cost of buying extra domains.
- More time spent setting campaigns up.
What to take from this
If you’re a newbie and are bit overwhelmed by having multiple domains and more costs etc., then yes you can get away with running everything from one master domain.
If you’re more serious about this business, your landing pages are assets to be protected and a small difference in conversion rate is important to your bottom line – diversify your domains.
But whoever you are – (pro tip) I wouldn’t recommend putting the word “scam” in your domain name.
NB: International Effects
I also ran an identical test in Brazil, with the same landers and same domains. It wasn’t until halfway through that I even considered that English domain names may not have the same influence over Portuguese speakers…. Duuuuh.
Conclusions from the preliminary results of that test: perhaps Brazilians don’t know the meaning of the word “scam” as the landers were all converting roughly equally.
I’d love to hear your thoughts on this or any other arguments I may have missed!
If you have any other ideas/assumptions/myths you’d like me to put to the test - hit me up!
Great case study Ruby Tunes!
I think this is a pretty extreme example ( using the word scam ).
When I tested this on PPV ages ago the difference was much smaller.
Originally Posted by bbrock32
Finally, this where I will direct people who always ask this 
Loving the myth buster series! Looking forward for more to come!
Nice study man!
There is definitely a big benefit in having branded domains, or even a style of domains (e.g android888.com download888.com whatsapp888.com). Some advertisers have directly hunted me down because they knew my domain style.
Interesting read Ruby Tunes! Thanks for the efforts 
Sweet case study, thanks.
Curious though, did you use entry pops on any of these landers? Domain names are a pretty visible factor in these JS alerts...
@imdutch - No I didn’t use entry pops. That’s a great thought though - that would have been a super interesting addition!
Great test! It's really good to have solid data for this, as it's one of the oldest questions in AM.
I wonder how the results differ on web traffic? My guess is that the effect would be even smaller...
It's great to see someone finally getting around to testing this, but as you mentioned, the effect of running in Brazil might compromise the findings. Though it would be great to use this as a baseline against which to measure an English speaking country with exactly the same test. My guess is that the effects would be further amplified by the use of entry pops.
Ahhh thats awesome my man ruby haha, this is great. I've always purposely picked domains that are more in line with the niche, it definitely seems more of a positive to do so
Originally Posted by mcsteve
Originally Posted by Ruby Tunes
Love the Myth Busters concept mate!
We need to come up with some more myths for Ruby Tunes to bust (or not)! Suggestions?
Originally Posted by caurmen
I love this idea as a start to a series of great, empirically tested posts!
Very interesting indeed.
Even more, the CTR was almost the same on all, but the CR was different. I wonder if people noticed the domain and still clicked but then they wouldn't install the app because it looked like a SCAM.
This Myth busters threads should have their own section!
I have one reverse-engineering question related to it. Why would anyone add "Scam" inside the domain?
How come it would benefit you having in it?
Thanks for the case study mate.
Not sure if this has been mentioned but we use really generic and short URLs that we add subdomains too. It gives you a ton of creative freedom without having to constantly purchase, setup and provision new domains.
Depending on what you do you may need to cycle through them periodically.
Depending on what you do you may need to cycle through them periodically.
If you make the subdomain long enough the generic part will not even show up in intro pops and the browser header bar.
Especially for intro pops always custom domains because the beginning of the domain name will be spelled out ...
Originally Posted by aimisaudi
The reason I used the word scam was to compare the results with an extreme end of the scale. E.g. if there was no difference in performance with domains containing “scam” - it would be obvious that people really don’t pay any attention to a domain name. It wasn’t as a competitor as such, it was as a comparison
1 question. Should I have targeted niche domain for my
so lander (android-store.com) --> tracker (android-tracker.com) --> offer ???
Originally Posted by ericchuawc
In this case it does not matter, because it's only used for redirects, the user might see it for a part of a second if they have super slow internet. But usually they will not. The only one that's relevant is your lander domain name.
And the flow is more like:
traffic source --> tracker [only redirect] (android-tracker.com) --> lander [domain visible] (android-store.com) --> tracker [only redirect] (android-tracker.com) --> offer
By having access to the overall domain, how would one be able to get all the landers?
Aren't your landers still protected since they wouldn't know the specifics?
I think the scam domain's results are pretty obvious but there was a 13.8776% increase between 4e7g383.com and android-tools-today.com and that is not bad for a $10 one-time purchase. If you are not doing this then you are lazy AF.
Originally Posted by hinkfo
I think your test is flawed, as you don't really run a "good" domain against the others.
Don't you guys find it weird to expect people to convert on on a domain like "android-top-tools.com", it just sounds super catchy and untrustworthy. Still, it seem to have worked decently, maybe due to the traffic beeing mobile where the URL isn't fully visible.
It would have been interesting to test a domain like "repairandroid.com" or "antivirus.co.uk" and then measure the impact.
That beeing said, you did a test and I didn't, so props to you. But in terms of the results, I don't think they're worth anything
Considering generic user 's behavior, I think most of the users, even if they are on a lander where the domain is for sure visible won't even notice the domain name (even though I'd have my doubts).
I wonder if anyone has tested with generic domains like xvzertical.site for landing pages, also what would be the effects of having geo targeted TLDs for landers. Regarding security concerns maybe it's not something to worry of, taking in consideration that spy tools are there always doing their job and copyright is something to dream of in AM.
@platinium
I have tested with generic domain names and the difference in conversion was minimal to none.
So in short, start optimising your domain once you are doing enough volume where a 2-3% increase will make a substantial difference.
Home > Affiliate Marketing Forum >