How Amazon uses personalisation to get more sales

Feb 4, 2020 | CRO, UX

Ask anyone working in digital for their top five eCommerce websites, and chances are Amazon will feature heavily. Considered one of the world’s tech giants alongside Google, Apple and Facebook, it started as an online marketplace for books but has expanded to not only sell physical goods, it has added grocery, cloud services and also produces content for its Amazon Prime entertainment channel.

I have been an Amazon member since 1999 and initially used their service to purchase books, CDs and DVDs. They were cheap, and their range was vast and with their increased range I dread to think how much money I have spent with them since those early days. It’s worth noting they do particularly well with me around birthdays and Christmas, as not only can I purchase the products I want to buy, I can get a good idea of what my family and friends want (from their wishlist), and Amazon will wrap the presents up (I’m terrible at wrapping), and deliver them. They recognise that their market is not simply for the individual’s own needs but for the individual looking for ideas for others.

From a digital perspective, Amazon have led the way in many areas. We mentioned wish lists above, they also adopted reviews very early on, social proof, which is so important in conversion, whether that be physical products or services. Their checkout process, which is vital in any eCommerce design where upto 70% of carts are abandoned, is quick, simple and effective. Another area they do well is personalisation, which is what I wanted to take a look at here.

Have a look at the image below, above the fold on Amazon when no customer is signed in to their account.

Let’s look at a few things on this example as even without any personalisation information, you can get an understanding of some of Amazon’s thought processes.

Firstly, if your mouse goes anywhere near the top of the page it is encouraging you to sign in or sign up as a new customer. You can also see that further down they encourage you to sign in “for your best experience” – ie personalised! Under the main navigation is a rotating banner which focuses mainly on Amazon’s own products; Prime TV series, Kindle e-Readers, Fire Entertainment products, but also Amazon’s offer of a 30-day trial of their Prime service. Just below the fold they show examples of films you get included in a Prime subscription, so it’s clear getting people to sign up not only as a customer, but as a Prime customer is important. Note also, although Amazon knows nothing about this customer, with it being early February, they may be looking for ideas for Valentine’s Day, so this features early on the page.

Now compare to when I have signed in to my own Amazon account.

I should say at this point that I am a paid up Amazon Prime member, so I don’t see the offers of a free Amazon Prime trial. However, on the rotating banner, Amazon is offering me two services I don’t currently use, Amazon Music, and Prime Now which is their grocery delivery service. In the boxes underneath, Amazon has learned that my shopping preferences include sports equipment, electronics and computer accessories. Without scanning back through my shopping history, I’d say that’s pretty accurate. The centre box is not an example of my Amazon use, but my account is connected to my wife’s account and she has been watching the featured programme (which is an Amazon exclusive), so it’s encouraging me/us to watch more of it. And the final box shows a hand-held over which I looked at but didn’t buy a few days ago. Amazon recognises that I might be researching and gives an easy route to find the products I had looked at before.

Further down the page, there are more examples of Amazon learning what I like and therefore what I am more likely to buy.

Ok, so I haven’t been looking for a Smart LED Projector, but from my searches for electronic things, it’s a possibility. And of course, it’s on special offer (ie it’s a bargain, and it might not be tomorrow or next week).

But notice the little Ad feedback link below the advert…I can give feedback on the relevancy of the advert to me, and if I feel it is appropriate to be shown on my Amazon page. This is an example of Amazon testing things out not only with me but on the platform as a whole.

Underneath, the gift finder helps me to find gifts via a wide range of categories…age, sex, relationship etc… Amazon are making it easy for me to find gifts if I’m struggling for ideas.

These items are on my own Amazon wishlist, so an easy personalisation. By putting these items in front of me with a “treat yourself” type headline, it may well tempt me to add something to whatever I came on to the site to order.

Now compare to the items above. On my wish list were books such as sports or politicial autobiographies, so Amazon are putting similar types of books in front of me. Even if I don’t purchase these, I sometimes spot an item that I wasn’t aware of so I add them to my wishlist for later.

Interestingly, this also brings up where Amazon don’t always get things quite right. You will see the Disney Frozen Annual which has come up, I assume because I have purchased Disney Frozen merchandise for one my younger relatives! I would be interested to know if Amazon can filter out items I have indicated are gifts, either by selecting them from someone else’s wishlist, or by selecting their wrapping services. This suggests that might not be the case and an area for improvement. Of course, I could be on the site to find some gift ideas for that same relative…!

Any sales channel likes customers guaranteed to come back on a regular basis to purchase. In the examples here, Amazon have picked up things I have bought and compared them statistically to items purchased more than once by the same customer. I’m not sure why someone would be a golf training aid more than once (maybe they are a coach), but the Block Salt may be of interest to me as it’s used in a household water softener. So why would I need to go elsewhere?!

Amazon have built their sales machine based not on guesswork, but on continual testing of ideas, making small changes over time so as to make improvements without damaging existing ideas that are already working. In your own business, and assuming your website visitors have given you permission to use cookies for personalisation, what examples could you take from Amazon for your own website…?


  • Customer sign-in
  • Wishlists
  • Browsing or order history
  • Product or Service suggestions
  • “Add-on” or complementary products
  • Regular purchase ideas
  • Gift ideas