We were happy to get media exposure for Fixel, the innovative online sneaker store that is using Artificial Intelligence to power their personalised shopping experience. Read about the latest developments in online retailing, as published on Bizcommunity.
The concept of hyper-personalisation has been around for a while now, and getting this right in online retail is key to a good customer experience, but are we there yet?
Amazon bills itself as ‘The Everything Store’, but rather than asking customers to digitally walk through isles showing everything they have on offer, they generate nearly 40% of their sales from their recommendations, which is now the main feature of their homepage. Walmart, its main competitor, is doing its best to catch up, but they face the almost impossible challenge of trying to replicate this in their bricks-and-mortar store environments.
Personalisation is key to a good customer experience and if the experience is better online, that’s where the growth is going to happen. PWC found that consumers will pay up to a 16% premium in exchange for a good customer experience. And what makes for a better CX than for a customer to feel heard and seen?
Levelling up personalisation
Personalisation at the level I have suggested though, goes beyond the common ‘customers who bought this… also bought this’ info box inserted at the bottom of a product listing.
In South Africa, Pick n Pay offers a level of personalisation through their Smart Shopper card. Based on customer purchasing, they give each customer different discount vouchers for the items that they actually buy regularly, and on products similar to the ones they have purchased. This saves customers having to page through their ‘specials’ sheets at the front door and gives a taste of personalisation.
But for a fully personalised shopping experience, you need to make the entire customer journey catered to the individual. Think of a Facebook or Instagram feed, no two are the same. The entire online store should rearrange itself according to your personal style, taste and needs.
Imagine if Pick n Pay arranged an aisle for you each time you walked into the store and only filled it with the products that met your unique preferences and needs at that given moment. They could give a truly personalised shopping experience by understanding what products you buy and when, but also give you discounts on the products you took off the shelf last time but then put back, and by removing the products you walked past and completely ignored.
When it comes to fashion, both The Yes and its more established competitor Stitch Fix are clear leaders in the United States clothing and footwear space. Both take advantage of recent advancements in machine learning and AI to power algorithms that work to understand a customer’s unique style, and then make suggestions and recommendations.
Stitch Fix is so confident in their suggestions, that they send their customers clothing that they haven’t purchased or haven’t even seen, and the customer only pays for what they keep. The machine learning recommendations have proved to be so accurate, that the company, founded only 10 years ago, does R30bn in annual sales – more than most of the listed clothing retailers in South Africa who have hundreds of stores and decades of experience.
Truly catering to your customer
We are starting to get there in South Africa. The first step is to get an initial understanding of a customer’s preferences by asking some basic questions upfront when they sign up for an online store. The next and far more complicated step is to assess the customer’s behaviour and how they interact with each product.
By gaining insight into what they stop to look at, how long they look at it, and what they scroll past and ignore, amongst many other behavioural attributes, you can start to truly understand what the customer wants and provide a very personal shopping experience.
The data and algorithms that power this experience create search results that drive sky-high interaction levels, low bounce rates and high conversion rates.
The machine measures how many times there is an accurate prediction of what the customer wants to see based on their interaction with the search results. If a customer is pleased with their selection, the machine ‘pats itself on the back’. If the customer ignores the recommendation and goes to look at something else, the machine ‘has some hard words’ with itself and tries something new next time.
Customers like not being overwhelmed. They like being an individual in a generic and faceless online world.
An April report published by PayU, titled The Next Frontier, indicates that there has been an increase of 35% in mobile transactions in South Africa since 2019. According to the report, South Africa’s internet penetration is 56%, and e-commerce sits at 37%, which highlights a significant opportunity for growth.
If we look at the data from our customer surveys, we see that once a customer is comfortable shopping online, they are not going back. So let’s make them comfortable.
First published on Bizcommunity.