Un-ique

  • 28 January 2011
  • 9 replies
  • 1 view

Would love to get a unique url to one of the variations… could be useful for showing people that you want to make sure you know exactly where you are sending them- could be useful for getting more traffic to a specific variation to test it more quickly


9 replies

I had the same problem… YOU CAN DO THIS. Just go to www.yoururl.com/a.html > www.yoururl.com/b.html > etc. It does NOT track leads when you use this page and it does NOT track visitors. Cheers!

Very neat trick. I love it. Takes forever to load but it’s a start. And what a bonus if this tracked leads and visitors (or better yet, gave an option if you wanted to “track visitors and leads with variant urls?”

Hey Adam and Anthony (thanks for jumping in Anthony) - just so you are aware, this trick (adding a.html or b.html etc to the end of the URL) is a bit of an undocumented feature for previewing specific variants and is not intended to be used for driving traffic (we may change how sharing/previewing works in the future at which point those links might stop working).

That being said – we will make sure we do our best to update all relevant threads if this ever changes. Adam - if you are looking to increase traffic to a specific variant to test it more quickly, I’d recommend just assigning it a higher traffic weight percentage than the other variants in the test. This will let you continue to gather stats and results while still using the same URL - will that accomplish the same thing for you?

@Carter: The only way to show a specific variant to your client for approval is the a.html method. If you are to remove this feature it would be very helpful to give a public URL that could be sent out for testing/approval. I often make 5 different variants all which have to be approved by someone else… without this method it is impossible.

@Adam: it actually isn’t slower to hit that link unbounce servers have been going a little slow today.

Hey Anthony – we definitely won’t remove the unique URL method of sharing individual variants without introducing a new method that lets you accomplish the same thing, so no worries!

We have noticed the slowness as well and will be making some optimizations tonight and tomorrow to speed things up again.

not necessarily. say i wanted to test the word BORROW vs the word RENT. I could set up an adwords campaign to test a hypothesis on one variation. So say I set up a campaign for RENT, upping the percentage for the corresponding variant on unbounce would HELP but the link would still be random and THAT is what is nice to have control over… there are other examples where I might want to remove the randomness as well.

I see, in that case I’d probably recommend duplicating your page (not just a variant) and using a different URL for each adwords campaign – that way you can properly measure the results of the whole funnel for each hypothesis independent from each other. If you really want to use the same page, but have 100% of the traffic targeting a specific variant then you could always just promote that variant to champion and discard the other challengers until you’re ready to bring them back into the experiment. We do have some plans in the future to overhaul our testing and experiments section so that it’s easier to start and stop tests and view test history etc. - maybe that will help here too.

I just used this method to send a test of variant A and B, but it tracked a visitor and conversion on my form. I was under the thinking that conversions using this method would not be tracked. Any insight on this would be appreciated. It’s not a big deal for this one campaign but definitely will affect future sharing of pages for approval/insight from colleagues.

Hi Luke, conversions will be counted and lead submissions will be saved - visitor stats will not be recorded though. Sorry if I was unclear earlier.

Reply