Clients | 11 Feb, 2018
Finding Fashion that Gets You
Your Flipboard magazine picks out the top stories that matter to you every day. Your Instagram feed auto-magically fills with funny animal posts so you get to start every day with a smile. Netflix knows just the kind of shows you're into.

When everything around you is perfectly curated to your liking, why can't finding fashion that suits your sentiments be just as easy to find?

Sounds so perfectly logical, doesn't it?

It did, to us.

Enter Pi — our unique personal recommendation engine that acts as an online shopping advisor and helps customers find products that are suited to them.
Pixibo has always been about simplifying those small, seemingly inconsequential things that make or break an online shopping experience. We realized solving the size and fit dilemma was only one part of the puzzle. Improving the product discovery experience involved understanding who we were talking to, at a personal level.

Personalized product recommendations directly have an impact on online shopping and conversion rates. Customers who click on a recommended product have a 70% higher purchase rate (at 10.5% versus 6.2%) within that session.

Our recommendation engine is built to curate a thoroughly personalized set of product suggestions to every shopper on the site. All shoppers are asked a set of questions that help indicate their style preferences. This information is combined with existing data about the shopper's past purchases, returns etc. to present them with products that are rightly suited to their likes and dislikes. The algorithm also incorporates a 'did not like' feature which eliminates similar products from future searches on the site.
Customers who click on a recommended product have a 70% higher purchase rate (at 10.5% versus 6.2%) within that session.
We built Pi because of the obvious drawbacks of most recommendation services. Basic recommendation engines require a whole lot of data about a customer in order to offer them worthwhile product suggestions that are geared towards conversion. And unless you're Amazon, it's hard to come by such data. In the absence of such in-depth data about shopper preferences, most mature recommendation engines have limited information such as age, geography, and gender. As a result, their product recommendations are far from exact and lead to low conversion rates.


Pi, on the other hand, removes any friction that exists in the path to purchase by showing shoppers the kind of products they are most likely to buy. A scalable personal assistant to every shopper on the site, it custom curates a personalized mix of outfits in the right sizes. What you get is a personalized shopping experience designed to mimic the attention and service you would get from a sales assistant at a store. Except, it's designed to know you even better.


We're all set to launch Pi on Indonesia's www.mapemall.com. Watch this space for more.
Kyu Lim is the Design and UX lead at Pixibo. She is often found empathising with users and creating digital experiences that don't involve big flashing buttons.