"Mad Libs" Style Form Increases Conversion 25-40%
- Thread starter stan100
- Start date
fantastic idea- stats are worthless
tons of data has shown call to action button SIZE is important- am I really the only one who noticed the call to action a) changed size by about a factor of 3, and b) changed to more understandable messaging
"send" (WTF? Try "submit" idiots) to - "send this message"
edit: - you also get a lot more people off the fence by lowering the barrier of entry- they can just bang in name/email and KNOW the dealer is being sent which vehicle/listing it was in response to. This may have been happening in the first example however many users don't want to take the time to 'explain' the situation in some piddly little 'comments' box.
tons of data has shown call to action button SIZE is important- am I really the only one who noticed the call to action a) changed size by about a factor of 3, and b) changed to more understandable messaging
"send" (WTF? Try "submit" idiots) to - "send this message"
edit: - you also get a lot more people off the fence by lowering the barrier of entry- they can just bang in name/email and KNOW the dealer is being sent which vehicle/listing it was in response to. This may have been happening in the first example however many users don't want to take the time to 'explain' the situation in some piddly little 'comments' box.
It would probably even work better if more sites used this style. I bet some people would have converted (with the old form, for example), but instead didn't because they got confused with the new form. Confusion causes friction; interest provokes action. So, it's really a balance you have to find.
The mad libs stuff is cool and I'm sure contributes to the increased conversions, but even without mad libs, I think when people see forms like the one on the left, it triggers "damn, another form to fill out" because they can easily count the different fields they will have to fill out. Then when they submit the form, in their mind (hypothetically), it would go into a stack of 1000 other forms and someone will eventually, maybe get to it.
The form on the right looks more like a message, in their mind (hypothetically) someone on the other end gets a text whenever it is submitted. They are also typing it as they read, so they're actively engaged as opposed to the traditional format, where most people know what is supposed to go into the fields without even looking at it.
The form on the right looks more like a message, in their mind (hypothetically) someone on the other end gets a text whenever it is submitted. They are also typing it as they read, so they're actively engaged as opposed to the traditional format, where most people know what is supposed to go into the fields without even looking at it.
I think it also triggers the need to complete sentences. We have a natural tendency in communication to do this. When someone is talking to you your brain already forms a path of where you think that conversation is going to lead. The path is constantly reforming as the conversation evolves. So this form vs the traditional one is a sort of neurological call to action that compels you to complete the form and submit.
Exactly, this split test was flawed from the start. I like the idea but damn why not keep every other variable the same so the results mean something. This has to be tested again before I would trust that the madlib form would really win if everything else was the same. Having product, pricing etc on one form and not the other can easily be the reason for the increased conversion rate.
I'd presume those are smart enough to look for field names but this is a valid question - I'd love to know too
ok, so i want to say that im actually working with local car delearships and TONIGHT im going to make a split test with this
Of course ill be reporting back whenever i have the results.
YES, every other variable will be leave as it is, so my results are going to reflect a real split test.
Good luck bros
Of course ill be reporting back whenever i have the results.
YES, every other variable will be leave as it is, so my results are going to reflect a real split test.
Good luck bros
I respectfully disagree. My reason for making this thread comes from the others before it and the ones that come after. Can't wait to hear about the live WF test :338: