top | item 2327110

How to increase signups by 50% using “popup forms”

43 points| sparshgupta | 15 years ago |visualwebsiteoptimizer.com | reply

38 comments

order
[+] patio11|15 years ago|reply
As always, test test test. I tried this for BCC, repeatedly, and never got it to work.

Also as always: do not assume that you're seeing the result of any one element unless you are testing that element in isolation. You might think the lightbox element is the most prominent aspect, but the call to action also changed from "Start Optimizing Your Website" to "Create A/B Test ->"

You see that arrow there? That arrow, tested in isolation, has been worth 10% to me before in some circumstances. Yeah, it blows my mind, too. And there were many other things that changed.

[+] paraschopra|15 years ago|reply
Interesting, yes. Agreed that there are too many variables which we tweaked. But this was a two stage A/B test so if you remove this popup box and move back to traditional signup form, we saw conversions drop back to normal.

May be it was the arrow icon that did the trick but I am not complaining since we saw such good increase in conversions! ;)

[+] ihumanable|15 years ago|reply
I can't agree more with this. We've run A/B Testing on this kind of modal dialog and we did not see an increase in conversion. In the article too many variables changed at once.

Changing the "Watch the Video" button to a "Start Optimizing" button is a huge change in and of itself.

[+] koolaidavoider|15 years ago|reply
Not only that, but do not assume that what works for one audience automatically translates (which is when testing helps). Different websites/products should target differently.

So even if this worked for one of your sites, it may not be the best thing to do with another.

[+] petenixey|15 years ago|reply
This reminds me of the draw of the "next article" dialogue that slides in on the bottom right of NYTimes articles - it makes for a very compelling click.

The (not so simple) key to these UX tricks is to do them in a way that they are a service to the user which also happens to be one of the points of Jakob Nielson's latest alertbox: http://www.useit.com/alertbox/hated-design.html.

It would be interesting to see how much of the value of this you could get without an overlay and by adding eye-catching dynamism to the signup dialogue - sliding the middle of the page away to reveal it for instance.

EDIT It would actually be interesting to see how much of this benefit would come from simply having the signup form itself on the frontpage - all previous designs required a click to reach it

[+] petercooper|15 years ago|reply
Why not go a step further and have the signup right there on the homepage? :-) Tumblr does it and I've been getting killer conversion rates on Ruby Weekly and JavaScript Weekly doing it too.

If your form is too big, I like Focus.com's approach of having the simplest part of the form on the homepage which then leads you into the full deal. I'd love to know how that's performing for them.

[+] paraschopra|15 years ago|reply
Yep, that's exactly what we are planning to test next. There are SO many things to test. Exciting times..
[+] BenSS|15 years ago|reply
Everyone thinks these things are hot stuff lately. I hate them because I'm there to read the content I was linked to. I won't have an opinion of if I'm even interested in subscribing/donating/etc until AFTER I've finished reading it so I don't even read them. I just hit close.

Essentially, they piss off first-time visitors to increase repeat visitor signups.

[+] paraschopra|15 years ago|reply
I think you didn't read the case study. We don't popup the signup form as soon as homepage loads, rather we popup when people click on Start Optimizing button on the homepage.
[+] kiiski|15 years ago|reply
Note that this form pops up only after the user clicks on a button:

"Implementation wise, here is what we did: on homepage, whenever a visitor clicks on Start Optimizing Now button, we show a “popup signup form” which has only two fields (username and password)"

[+] jneal|15 years ago|reply
I've used similar functionality on a real estate website, allowing people to see that the information they want is behind the registration form. Previously, the client averaged 1-2 signups a day. After the change, they started seeing 6-7 per day. So this definitely can work - but as with every other user interface component you should use A/B testing and see if it works for you before committing to it 100%.
[+] Tichy|15 years ago|reply
Nice and all, but I have started to wonder if the A/B-Testing mania is just the new SEO mania, resulting in just as much snake oil in the end.
[+] epe|15 years ago|reply
I think the snake-oil nature of (much of) SEO stems from the fact that it's difficult or impossible to measure in an accurate, unambiguous way. Since A/B testing includes measurement by its very nature, I suspect it will be more resistant to snake oil.
[+] paraschopra|15 years ago|reply
Can you please elaborate how is it snake oil? I am genuinely curious since I run an A/B testing company.
[+] lothar|15 years ago|reply
I think popup forms are annoying and spammy. I'll never use them on a website of mine.

Using them might result in more email signups, but I think the response to your email blasts will be worse off and result in a lot of immediate "unsubscribes".

Popups are a trend I hope goes the way of blinking text.

[+] chaz|15 years ago|reply
> We used our new revenue tracking feature to see impact of “popup forms” on actual $$ made, however it is still to early to say anything.

I think this last sentence is key. Raise a barrier and of course the number will go up, but average value per signup will go down. This might be fine for freemium or very good email marketing, but it may also have a negative effect. Only further testing will reveal the complete funnel around total visitor value.

[+] rexf|15 years ago|reply
Is it considered acceptable practice to not include a 2nd password verification text field?

I can see that there's less friction in the process (1 less field to fill out). Perhaps since this is a specific lightbox call to action instance and not the traditional register page, there is no password confirmation text field (meaning the full fledged register page would have a pw confirm field).

[+] ernestipark|15 years ago|reply
I have not used a second password verification text field in some of my sites. While I have not tested it, I prefer it. Instead, if someone thinks they have mistyped their password, they might just retype the whole thing. Or, I give them an option to "show" or "hide" the password which toggles between password dots and normal characters.
[+] mattcurry|15 years ago|reply
I found that when I removed the second password field users thought they were signing instead of registering. This lead to users creating multiple accounts and panic-y emails about lost data.
[+] spontaneus|15 years ago|reply
Having your call to action button below the fold is really bad practice. It's no surprise that your conversions increased when you moved the button up. Also, I would suggest putting your registration form directly on the homepage instead of requiring a person to click a button to see it.
[+] richcollins|15 years ago|reply
I'd wager that a landing page that just had the contents of the pop up would do even better.
[+] paraschopra|15 years ago|reply
Yep, that's what we are going to test next.
[+] jimlast|15 years ago|reply
We converted ShrinkingApp.com to modal signup forms 6 months ago or so and saw a slight increase in conversions, nothing too dramatic. I can see how it would work from a simplicity standpoint (less page loads, quicker, etc.).
[+] rheide|15 years ago|reply
This is ridiculous! Of course more people will sign up if you make them only fill out two things and force the popup on them, but it completely devoids the concept of 'signup' of its meaning. What you should care about is how many people sign up and use your service. If I was presented with a signup popup like that I would just input a fake e-mail address, try it out once and never come back. That's not what a signup is supposed to be.