Fundamental Guidelines Of E-Commerce Checkout Design | Hacker News Reader

Fundamental Guidelines Of E-Commerce Checkout Design

smashingmagazine.com

70 pointsgulbrandr16 comments

Is there raw data for this available? I would like to see survey responses and footage of the testers as well.

Great guide but I really think this area needs a lot of innovation, the checkout process isn't nearly as optimized as it should be. It's 2011 and you still need to fill out the form every single time you want to buy anything.

I think we need to aim for the one-click-to-buy dream to get the checkout process getting it's usability and conversion it has the potential for. Apple's App Store is a great exampel of how easy we can make this come true. Two clicks and a password and boom you have bought the app. I want the same standard on all e-commerce sites, possibly with Facebook Credits/Facebook Connect, so whenever I'm logged in the purchase will take 1-3 clicks and no damn forms to fill out.

Maybe I'm the exception, but I don't want checking out to be too easy:

- I like to type my credit card each time, I don't want sites storing it

- I don't want to accidentally click things and buy things I don't want (I click things a lot, just to 'see what they do')

- I want the buying process to require _some_ effort, so I only end up buying things I really want.

Of course it will require some efforts. I think the iPhone App Store purchase process handle it very nice, it requires enought effort but still makes it smooth and easy.

Usability and the most simple solution will always win in the long run.

I think we need to aim for the one-click-to-buy dream

Not while Amazon holds the patent! Though I guess 2 or 3 clicks is still an improvement.

In working with thousands of small businesses I've found many more exceptions than rules. I think the key to checkout is for businesses to optimize for the most common customer of their store. This may go against common wisdom for general stores but if you optimize for your customers you'll reduce the checkout workload. Less work during checkout has a direct correlation to higher completion/conversion.

Interesting but there are a lot of things I would not suggest. For example, the article suggests that expiration month on credit cards be listed "exactly as it appears on the card" while failing to mention that there is no standard. The author suggests you do NOT include the name of the month next to the number but many cards use the name of the month so listing expiration as: 01-Jan, 02-Feb makes a lot of customers happy. I would read this article as great suggestions but it doesn't mean the advice applies to your store perfectly.

I have never seen month represented using the month name - at least not in Canada.

Don't know what its like in other areas of the world. Perhaps we just need to use some critical thought after reading and interpret the points based on our markets.

I mean, it isn't a hard an fast rule - they are just giving some suggestions - and I think they were pretty spot on.

I just looked through my wallet and Amex, Visa, MasterCard, and Discover all have the same date format, and for MasterCard three are from separate issuing banks, one of which is a debit card.

Not questioning you, but what's the rule? Do you have any examples? Is it a function of the processor or the bank?

Sites like WalMart and Zappos are mentioned in the article for doing things that are non-ideal, according to Smashing's testing. It seems pretty inconceivable for sites with such large revenues to not have done their own internal usability testing. I wonder what led these sites to their current usability/design decisions.

Online is a different world. Don't assume those who "get" brick and mortar can also "get" online. When walmart first chose to go online they realized this might be a problem and hired a group out of California (this was during the dot com bubble) to come in and do it. Those guys didn't do a very good job and there was a bunch of noise being made by internal people about "we're good enough to do this! We don't need some fancy California people!" so it came back in house. Well they weren't good enough (to be fair, the "walmart online" people weren't either).

Though, for me I always figured Walmart.com's biggest problem would be that they couldn't take advantage of the internet tax rules: they have a presence in every state afaik.

You seem to have a bit of understanding of what happened internally at walmart with walmart.com in the early days. Walmart.com is still based out of California (Menlo Park I believe) even though the company was brought back into the Walmart Inc. many years ago.

The original "walmart online" folks weren't web developers (Visual Basic coders) and really had a tough time understanding the request/response nature of the web (although they had no UX or QA people either, which led to other issues).

The Arkansas/California issues you talked about also stemmed from jealousy. There were stories of the associates in CA having pool tables, foosball tables, catered food, etc. Meanwhile, in AR the associates had extremely old PCs, old software, required to work long hours, etc.

I understand your premise. I'd expect on online-only retailer to have better intelligence on revenue optimization than a B&M giant like Wal-Mart. However, this seems like it's becoming less and less of a norm.

I know of several large B&M companies who also do business online, and they have their own online units, which certainly includes UX people. I couldn't imagine Wal-Mart neglecting WalMart.com. The UX issues of walmart.com brought forth in this article weren't that big of a deal. If they weren't paying closer attention, I'd expect them to have much more substantial problems.

Being 'able' to do something is far more than technical chops. I'd think for a company the size of Walmart there's so much internal political stuff that has to be dealt with - competing departments, etc - that regardless of what people are implementing the tech, what is directed to be implemented is a poor compromise.

They mention using a single column, but what about displaying the address fields in similar way to how they're shown on paper, i.e. city, state zip in one line? Would this fall under one-column or two-column presentation?

Just to double down on one of their points about showing security around sensitive fields - one of the bread and butter things we do on optimization projects is test trust logos and trust logo placement.

Getting the best vs. worst positioning can drive as much as 15 - 20% on cart abandon - more details in article:

http://conversionvoodoo.com/blog/2010/07/proper-placement-of...

Reply on news.ycombinator.com