Our Brief Experiment with Big Commerce's Optimized One Page Checkout Has Ended, We've Returned to Our Old Checkout

Some of you may have noticed, but we recently ran an experiment using a different checkout than usual for the past week.  We did this because it 1) allowed us to offer amazon pay as a checkout option and had a direct link to paypal credit instead of indirect through paypal, 2) allowed us to use better analytics to better see and use our data, 3) promised to load faster, and 4) promised to increase our conversion rate by lowering abandoned checkouts.  Unfortunately this checkout proved to be less superior to our 1 page checkout app we use through intuit solutions and paypal.  Here's what went wrong.

First and most importantly, our conversion rate went down significantly, by 25% of what it has been in recently weeks.  Also because of the analytics options we could see where exactly people left the checkout and stopped being willing to buy the product from our website, which turned out to be the very 1st step of the checkout with over 50% leaving immediately.  We attribute this to a couple of factors: 

a) Optimized one page checkout looked different from our website.  It was basically pure white and looked very sterile and this was a bit jarring.  This was the most minor factor though.

b) It immediately had an account signup at the start of the checkout.  Of course the checkout as guest option is available, but it still forced users to enter an email to checkout as guest and we think that caused some confusion by making people think they had to have an account to checkout, which also leads into the 3rd point.

c) It asks you for an email separately from billing and shipping.  Separating the email from this section raises concern, like I mentioned in the last point, but also just being asked to enter your email on it's own seems strange compared to being part of a list of information needed to ship a product and send order and shipping confirmations.

Another thing that did not work out is that nobody really used amazon pay and also amazon pay created some accounting issues.  Paypal handles all of our checkouts, whether the customer checks out by entering their information on our site or clicking the complete through paypal button.  They give us the payment and tell us payment was received right away.  Amazon pay not only waits a few days before this notification, but we don't receive the money immediately as well.  As a result, we've removed this option, which is also not allowed outside of the optimized checkout page so it would have been removed regardless.

One thing the optimized checkout did do was load slightly faster on both desktop and mobile, but not enough that a person could tell, except perhaps on an older smart phone.  Unfortunately most of the promises of the optimized checkout did not hold up or they turned out to not be so good and we have switched back to our older better checkout.  The big downside for us is that we lose ecommerce tracking from google analytics and we have less data to work with, specifically the ability to see when people abandon carts or the checkout to try to determine the cause of it and improve our website, so we are forced to make less educated guesses again.  We also lose the direct link to paypal credit in the checkout, but the option is still available when completing the checkout through paypal's website as one of the payment options (aside form credit card and paypal wallet, I believe).

We experiment with features to improve customer experience from time to time, but sometimes they do not work out.  We've had many successes in the past, but this is the 1st one that has gone so wrong from our expectations.