Here’s something from a project I have been working on off-and-on for the past few months: Rewriting the entire shopping-cart system of the Collectorz.com site.
From the shop, via the upsell interstitial page and the name/address page, to the payment details page and finally the order confirmation page. I have written some new back-end PHP classes and am now rewriting all website pages based on those classes, making the pages more consistent, and more importantly, easier to maintain. Easy maintenance is essential, because I am tweaking these pages quite often.
I have already posted about my upsell interstitial page a while ago, which was the first page I based on the new back-end classes. This long Easter weekend I have completely rewritten the first page of the check-out, the shop page. Continue reading
At the SIC in July 2008, WinZip’s Edwin Siebesma told me that I should talk to the TrialPay people. I did and I liked their sales pitch.
Of course, like everyone else who first hears about TrialPay, I had fears of cannibalizing my existing sales by offering a TrialPay option on my website.
So in August 2008 I ran a simple A/B split test, testing just that:
- Group A: my regular product pages, no TrialPay option
- Group B: An extra “Get it Free” button, taking visitors to the TrialPay offer page. Just for our Standard edition (regular price $29.95). The button was a regular HTML button, just like our other “Add to Cart” buttons. Like so ->
And a week has gone by again. But now we’re finally seeing better results from the “B design”. Phew…
In terms of numbers of sales, it’s doing about the same as the old design. The profits for category B are slightly higher (about 3%), mainly because the average first purchase is still higher for the new design.
All in all, the new product pages are not a great success. But at least they are now performing well enough to switch to the new pages and start optimizing their details. I already have a couple of ideas I want to try:
- Headlines: Emotional / Fluffy / Benefit oriented vs Practical / Feature oriented
- Pricing: .95 vs rounded pricing
- Upsells: Move stuff like CD Delivery, Priority Support etc… to an interstitial upsell page, as opposed to including them in my “Recommended Buying Choices”
Here’s the results after another week of testing
(results of new design compared to the old design):
- Downloads: down 4.22%
- Sign ups: up 1.50%
- Sales: down 19,11%
- Profits: down 11,52%
- Average First Purchase Value: up 10%
Something’s not right in the new design, but it’s hard to determine what it is.
Yes, I know, that is because we changed too many things at once.
The loss of the testimonials on the top right of our new website design kept nagging me. So I discussed it with Mark, one of our web developers. And he came up with an idea to create a nice spot on the top right for one hot testimonial:
Last week, I told you about the A/B split test I started, comparing my old website design vs a radical new design. I was planning to let it run for 2 weeks, unless the results after 1 week would be dramatic. Well, they are:
These are the results after 7 days (A is the old design, B is the new design):
- Downloads: About the same
- Sign ups: 20 percent more for design B
- Sales: 19 percent less for design B
- Profits: 10 percent less for design B
- Average First Purchase Value: 10 percent higher for B
The design of our website is being changed almost continously. I am always trying to improve its conversion power, by making small optimizations to the layout, the design and the copy.
But this time I am implementing a more radical change: a completely new layout. In the next couple of weeks I will be testing the new layout vs the old layout using an A/B split test. That is, fifty percent of new visitors will be seeing the old layout, the other fifty percent will get the new layout. Then we’ll be measuring sign-ups, downloads, sales and profits in both groups.