A leading fashion marketplace shares findings from its eCommerce personalization testing
I am obsessed with electronic gadgets. So when an eCommerce store recently announced a sale, I was excited to find the next shiny gadget to add to my collection. As I explored the electronics section of the online store, I noticed that the recommendations were just a reflection of the products I’d already purchased in the past. After a few searches, I did not see anything exciting, but as I started digging deeper, I did find interesting niche products that I haven’t purchased before. The personalization was therefore restrictive and needlessly narrow.
This made me wonder – What level of personalization is best for an eCommerce business? What weightage should you give to an individual’s preferences, past purchases, and current context? How much exploration is ideal? There are many opinions on this matter, but one thing is clear — over-personalization can have a detrimental impact on business.
In today’s digital-first world, algorithmic 1:1 personalization is, without a doubt, a non-negotiable business strategy. It has proved to be a powerful tool to improve conversions, build brand perception, and boost revenue.
- 80% of shoppers are more likely to buy from a brand that provides personalized experiences. – Epsilon
- 89% of digital businesses are investing in personalization. – Forrester
- 80% of companies experienced a revenue uplift since implementing personalization. – Econsultancy
In a time when online shoppers are spoiled for choice, brands that are not tailoring experiences face the risk of losing their relevance. But there’s a catch. Too much personalization can come across as invasive to shoppers, create an echo chamber, and result in low conversion rates. So, how much is too much? The simple answer is — there is no silver bullet. You’d need to experiment and find what works for your business and markets. Let’s look at an example to understand this better.
Miinto, a leading European online fashion marketplace with 800,000 products, uses Algonomy’s all-in-one personalization software to deliver hyper-personalized experiences across the commerce lifecycle and across the touchpoints of recommendations, content, search, and browse.
In a recent conversation, they shared the results of their tests. The company carried out a series of A/B tests for one of their markets — gradually increasing the personalization level of the product assortment on category pages — and monitored the impact on:
- Revenue Per Visit (RPV)
- Conversion Rate (CR)
- Average Order Value (AOV)
In the first test, they increased the personalization from 25% to 50%. This led to a 5% uplift in RPV, which was quite significant. Since there was an improvement in all KPIs, they were curious to see how far they could go.
Subsequently, they increased the level to 60%, 65%, and finally 70%. They found that at 70%, there was a sudden decline in all KPIs. This is perhaps the point where personalization became more creepy than helpful to shoppers.
65% was the sweet spot, where they got the highest uplift in RPV at 11%. See the figure below. From these results, we can infer that while personalization is crucial, there is a fine line between being relevant and invasive.
More businesses need to embrace an experimentation culture and find the personalization level that works best for their customers. After all, every vertical and market is different. Further, it is important to remember that for a shopper, relevance is certainly important, but so is the joy of discovering new products and categories. Therefore, businesses must ensure the right blend of enabling hyper-personalization and allowing for discovery and exploration.
In the digital-first era, it is imperative to possess the technological capability that allows for rapid personalization tests — an area where many retailers are still lagging. Brands also need to move beyond segmentation and treat each customer as an individual. The solution to this is to build real-time, individual shopper profiles and drive the optimum level of personalization for not only product recommendations, but also search, browse, content, and promotions. This is the key to growing and retaining your customer base, and accelerating growth in the intensely competitive eCommerce industry.
Curious to know more about Miinto’s customer experience journey and how their revenues dropped over 19% when they accidentally turned off Algonomy personalization? Watch the insightful session, from Algonomy Personalization Summit 2021, featuring Paloma Truong, Miinto’s Head of Customer Experience.
Read more on how to build end-to-end eCommerce personalization capabilities to address key commerce challenges.
The article was first published in ET Brand Equity