Digging Into Customer Data on Socks Site Turns 5x ROI

Ahua Apparel is a multi-brand Korean fashion and accessories retailer whose main product used to be novelty socks. A lot has changed since then, but the company slogan remains the same, “Ahua, are you ok?” Mr. Ahua himself is definitely “OK,” especially since he started using personalization to increase his average order value 19.6%.


Tight margins and difficult-to-manage customer data


Automated data analysis and fashion-optimized recommendations



Average Order Value increase


Return on Investment


“By analyzing customer preferences, I can easily find the most-clicked products, the shared attributes of the things added to the cart most often, the items with the most purchases and a lot more. That helps me learn more about customers and our products.”

Rosetta.ai case study


In 2019 shrinking profit margins on novelty Korean socks were leaving too little room for discounts. Mr. Ahua needed to find a new solution to increase profit and grow business.

He knew that increasing average order value is one way out of the tight margin trap and had heard that personalized marketing can raise AOV.

He searched around and found a local SaaS company specializing in personalized product recommendations for apparel websites. He reached out to them to ask about adding a recommendation engine to his website.
The Problem

Tight margins and difficult-to-manage customer data

Tight margins are normal for many online apparel merchants. Pressure to discount, rising returns, and all kinds of unforeseen costs can cut into profit and slow down growth. He also knew increasing AOV is all about making better connections with current customers and cross selling to them, a strategy used by some of the biggest players in ecommerce.

The key to this solution was right under his nose, in his customer data. But putting his data into an actionable business plan was challenging. Mr. Ahua concedes, “we had a lot of difficulty exporting our website data and analyzing it by ourselves.” It’s not an uncommon problem. Marketers all over the world have tons of data on hand, but find it difficult to manage
The Solution

Fashion-optimized recommendations, automated consumer data insights

The Rosetta.ai personalized marketing solution collects customer/product data for Mr. Ahua and machine learning figures out which other products individual shoppers want to see.

With recommendation engines installed on the landing page, category page and product details page, shoppers visiting Ahuaruok may come for custom socks, but they may add something else to their shopping cart now, perhaps a dress or a jumpsuit that goes with the socks.

The “You may also like” recommendation on the Ahuaruok product page.

With the customer data insights from the Rosetta.ai in hand, Mr. Ahua was soon able to see what was selling and what was not on the level of product attributes. To Mr. Ahua, customer insights are the most valuable thing, he even added “it’s not all about the money.” Now he views growing his business via cross selling much more favorably than through discounts alone.

WIth an understanding of the shared attributes of the things that get purchased most often Ahuaruok has a deeper understanding of its customers and products which lets him make more truly data-driven decisions.


Average Order Value increase


Return on Investment

While still in the 14-day trial period, Mr. Ahua noticed that average order value had jumped from 70 to 100 NT dollars per customer so he signed up for a full year subscription to the service.

Just 3 months later (by August of 2020) personalized marketing provided by Rosetta.ai was driving 20% of total purchases and 20% of total revenue per month.

By the end the year, AOV was 19.6% higher than before having the recommendation engine. The extra revenue translated into a 5x return on investment!

Enabling merchants to understand the language of fashion.

Get Started