Comments on: I Redesigned My Software. Users: Thrilled. Conversion Rates: Up. Sales: Unchanged.

By: Software Marketing Tweetables - 30 April 2012 | Smart Software Marketing

Software Marketing Tweetables - 30 April 2012 | Smart Software Marketing — Mon, 30 Apr 2012 10:10:22 +0000

[…] I Redesigned My Software. Users: Thrilled. Conversion Rates: Up. Sales: Unchanged. […]

By: Sergey Kornilov

Sergey Kornilov — Thu, 26 Apr 2012 05:07:12 +0000

I would keep testing it till statistical significance is reach.

On the other side – even is sales hasn’t changed users love new design and it’s going to pay for itself one way or another.

By: Web-based Startups: A Design Primer | the Intervals Blog by Pelago

Web-based Startups: A Design Primer | the Intervals Blog by Pelago — Mon, 23 Apr 2012 17:52:50 +0000

[…] or higher conversion rates, for that matter. We often times give design more power then we should, as this particular Web designer found out while A/B testing his new redesign. Apple did not succeed on it’s design alone, nor can we. We still have to invest time and […]

By: bad breath

bad breath — Sun, 22 Apr 2012 13:51:08 +0000

Very good blog you have here but I was wondering if you knew of any community forums that cover the same topics discussed in this article? I’d really like to be a part of community where I can get opinions from other knowledgeable people that share the same interest. If you have any suggestions, please let me know. Bless you!

By: Pit

Pit — Fri, 20 Apr 2012 14:31:25 +0000

New alternative to A/B or split testing is Myna (mynaweb.com). Myna is based on many years of research in decision theory. This testing method is used by big companies like Google, Microsoft or Yahoo!

By: Limo hire leeds

Limo hire leeds — Fri, 20 Apr 2012 11:20:27 +0000

…Links…

[…]I just added your blog to my blogroll, I hope you’d consider doing the same.[…]…

By: Iain

Iain — Fri, 20 Apr 2012 10:49:43 +0000

The small number of conversions makes statistical significance hard to achieve. The reasonable rule-of-thumb that “the square of half the difference needs to be at least as much as the number of tests*” to indicate 95% significance, means that for 26 conversions one would need to out-perform the other by 10 – so anything less that an 8/18 split either way can’t be relied on. For anything less than a few hundred tests – do the math – your “instinct” will get it wrong.

* Jason’s got a nice explanation/derivation for that here: http://blog.asmartbear.com/easy-statistics-for-adwords-ab-testing-and-hamsters.html

By: A revolutionary video game, a sketchnote on emotion in design and more | 21times

Fri, 20 Apr 2012 09:31:36 +0000

[…] I redesigned my software. Users: thrilled. Conversion rates: up. Sales: unchanged. […]

By: Bram Cohen

Bram Cohen — Fri, 20 Apr 2012 08:27:24 +0000

Umm, Patrick, why don’t you A/B test reducing the number of free trial bingo cards people can get? That would be easy to do, and is directly supported by your leading (only?) theory as to why sales are unchanged with the improved site.

By: Steve Moyer

Steve Moyer — Fri, 20 Apr 2012 01:48:51 +0000

Hmmm … so you’ve proven something you already know, but do you have any strategies for locating where you’re funnel is leaking? On a completely separate note, you’re dates are getting hard to read on the statistics page … can they be added to the mouse-over?

Thanks for sharing with us once again!