Amazon's Reverse Journey

Published: 02-Dec-2020

The Curious Case of Amazon's Homepage

I was scrolling through my Twitter feed mindlessly as usual when I saw this interesting question by Aditya.

It's interesting because you would expect a customer-centric company like Amazon to have fixed this by now. Surely they have all the data to personalize your home page - what kind of products you buy, when you buy, which brands or price ranges you prefer, etc. They should be able to use this to predict what I might need/want to buy and show them to me on the homepage.

But this is how my Amazon homepage looks.

Whatever little personalization is currently done like "top picks for you" or "inspired by your shopping trends" is either irrelevant or lost in this overload of pictures. So why has Amazon not fixed this? And will Amazon ever fix this?

My guess is Amazon will fix this....eventually. It's just not a focus area for them right now. Let me explain.

The Big Four

From an e-commerce shopping point of view, there are broadly 4 places Amazon can double-down on, to give a superior experience (see pic below). Amazon has been slowly working on improving each one but in the reverse order.

By the way, I really like wordplay. But I couldn't come up with a word for payments that ended with "y" and still made sense. Please help me.

Focus #1: Delivery

Amazon decided to improve the last step (delivery) first, and rightly so. Jeff Bezos realized very early on that if a product does not reach the customer quickly, probably nothing else matters. Getting a product to a customer's door a day after they clicked the buy button was (and still is) a strong way to deliver a magical experience to the customer. Probably so strong that they might be willing to overlook UX gaps in the ordering process. Because nowhere else (online) will they get something on their doorstep a day after they order. The excitement of buying a new thing is real and if the product can reach them during the post-ordering euphoric period, they experience delight.

So Bezos has burned through millions building a global fleet of 30,000 Amazon-branded delivery vehicles, 20,000 branded trailers, and 70 planes - numbers that are only going to grow. With this, Amazon has put the fundamental blocks in place to provide the quickest delivery possible to its customers.

A point of clarification before I proceed. When I say Amazon focused its money and energy on improving the delivery experience first, it does not mean that there was nothing done in the other steps. Amazon did have to string together the minimum so a customer could place an order (see pic below). But it did not bend over backwards to provide the best experience at the first 3 steps, at least not at first.

Focus #2: Payments

As the delivery machinery began to fall into place, Amazon moved to the step just prior to fulfilment - payments. Amazon Pay was built, iterated on, and made better. One-click payments — the magical experience in payments — was now possible. Bank integrations ensued and credit at the time of checkout was also made possible.

Focus #3: Advisory

With the building blocks of a 'better payment/credit experience' in place, Amazon has now begun to focus on strengthening advisory. While basic product information, pictures and ratings & reviews have been around from the beginning, Amazon has recently started listing demo videos and expert reviews in an effort to better help users in their purchase decision.

In the picture below, you can see videos from review sites like Digit and Times Now Digital apart from merchant videos.

This is only the beginning and I figure Amazon is working on a few more solutions to improve the experience at this step of the purchase journey. Problems like 'how to buy high-touch products' and 'how to weed out fake reviews' still persist. I do not know what the solutions to these problems are (and if they can even be solved online) but I will not be surprised if Amazon pilots some solutions soon.

However, this gets me to the answer for the question we had at the beginning of this post. It's only after improving the experience at the advisory step to a satisfactory level will Amazon begin to focus on improving the discovery step.

Discovery - the final frontier

Following the reverse journey to solve discovery last also makes sense from the point of view that Amazon will have enough data by then to get a decently functioning recommendation algorithm off the ground in the first attempt.

When Amazon actually does get down to improving discovery, we will see smaller changes at first - like a cleaner, more personalized home page. But as the accuracy of the recommendation algorithm grows, we could see some really exciting (and maybe scary) step changes. For instance, this HBR piece does a thought experiment on what Amazon could do IF their recommendation algorithm reaches a 95% accuracy level. It would then be (at scale) cheaper for Amazon to just ship the products to you (even before you buy) and let you return only those products you don't want. At a 95% accuracy level, you will end up returning only 1 in 20 items. Amazon will have captured your mind-share on 19 items even before you could begin thinking about the purchase and possibly buy them elsewhere.

Of course, this is all a guess from my end. Who knows, Amazon could clean up their homepage tomorrow and my entire theory will find itself in the bin. But till they do, we have to make do with a search bar. Cheers.

Sorry, one final holiday joke before I go.