The “Just Copy Amazon” Fallacy

I was talking to an executive at a mid-sized e-commerce business the other day, as he described his plan for looking at the industry leader to get “inspired” on how to redesign his checkout. He, like so many others, had Amazon in mind.

As the e-commerce godfather, selling for $61 billion and widely recognized for employing endless testing and optimization, Amazon is very often considered a “source of inspiration”. In fact, at all of the last three conferences I’ve presented at someone have asked if I thought they “should do [insert website element here] like Amazon does it..”

Here’s my take on it: Simply copying any industry leader, including Amazon, can be downright harmful to your conversion rates and should never be a substitute for testing yourself (usability, a/b, etc).

You can’t piggyback on Amazon’s testing by simply following their designs and implementations because:

1. Amazon’s business agenda isn’t yours. They make money in funny ways. Sometimes they sell at no profit, and “just” earn interest on the credit period from their vendors. They earn from ads on their pages. They sell anyone’s products, not just the ones they own or are in possession of. Therefore, their split test goal can be completely different from yours - the test version that makes Amazon the most money isn’t necessarily the one that will make you the most money.
2. “Everybody” already have prior experience with the site and the majority of visitors already have an existing Amazon account. This is likely not the case for your site. (In fact we found that Amazon’s 8 step checkout process for a new customer provides a sub-par checkout experience, likely so because the checkout is designed closely around the goal of being great for customers that already have an account).
3. Amazon’s chosen design is what works on average across their entire catalog. It’s OK that the design does not convert as well as it could in e.g. the apparel vertical, if it works better overall. Unless you are a mass merchant too, you may end up borrowing design elements that actually work poorly in your particular vertical.
4. Even Amazon mess up sometimes. For example, during our m-commerce study, where Amazon’s mobile site suffered from things such a sub-product pages and unclear primary buttons (see point 3 & 5 here), no explanation of password rules during account creation, no “sort by price” option for category lists, and more.
5. Your traffic sources aren’t the same. The level of persuasive design needed to convert a visitor is highly related to how “pre-sold” the customer is before even entering the site. Since you don’t have the same traffic sources as Amazon, your design will most likely need to persuade in a different ways than Amazon’s.
6. Customers have different experiences with and expectations to Amazon’s brand than they do to yours. For better or for worse, this makes a big impact in the level of trust and type of interactions they’re comfortable performing.

Of course, all of this goes for any industry leader, in any vertical – Amazon is simply the classic example. Also, do note that this is not to say that Amazon is doing anything wrong or that their true talent in the game of e-commerce shouldn’t be recognized. It is, however, to say that simply following the industry leader’s site is a perilous strategy because it doesn’t necessarily represent best practice and should never be done without thorough testing of its effects on your particular site and audience.

That said, looking at industry leaders to figure out new features and designs that could be interesting to test can certainly be meaningful as long as your own testing is at the core of such ventures. In other words, industry leaders can certainly be good sources of inspiration of what to test, but not necessarily what to implement.