What is True Fit?
True Fit Mobile
True fit is a technology startup company providing fit and style and recommendation engine support for online providers of clothing and footwear.
True fit boasts a database with over 10,000 brands, one million styles and more than 100 attributes used to classify styles as well as fit, body statistics, style affinity and demographic data from over 100 million global consumers.
Based on their fit and style algorithms, True fit integrates their recommendation engines into client websites where buying customers are able to compare clothing and shoes based on their own personal style habits, buying patterns, fit and purchase histories.
Reverse Engineering A Clothing and Footwear Recommender System
Understanding the layers and players in a complex system
Because True Fit is a third party provider or recommendation software as a service, their engine design must consider the needs of their customers, retailers and vendors selling both online and at brick and mortar stores, as well as the needs and considerationst of the end consumer buying shoes and clothes from their customers.
And it could be argued, that their own business needs, to build out products and services in a way which ensures continued growth in their sector, adds a third more complex layer of need to their design equation.
End User Analysis - The customers clients:
Who are your target users?
For True Fit’s Clients the target users are people shopping, both online and in brick and mortar stores for closthing and shoes.What are their key goals?
They key goals for the end consumer is to purchase clothing and shoes that fit well, look good, satisfie their own sense of style, and do not end up in a closet or drawer somewhere. (Studies show that less than 30% of average persons wardrobe are actually used regularly)
How can you help them accomplish those goals?
By suggestion clothing from a vendors line, that are similar in fit, style and appearance to clothing purchased by clients in the past, you are increasing the odds that the consumer will fins something which the like, will fit them comfortably and are in alignment with their sense of personal style.
Vendor Analysis - True Fit’s Clients:
Who are your target users? The clients of True fit, manufacturers, direct marketers and retailers (both on and offline).
What are their key goals? Our customers want to provide clothing and shoes to their clients which they will want to wear, feel good in, enjoying both the style and fit.
Knowing that for every online purchase are customer makes, they make 5 in-store, True Fit’s customers want to be sure that they increase customer satisfaction at first contact by improving the odds of providing the right look, fit and feel.
- How can you help them accomplish those goals?
By increasing the probability of a product meeting the end users personal objectives, True Fit can reduce returns, which are both costly, difficult to plan around and detrimental to client confidence in buying from a vendor online. So providing personalized recommendations based on an individual customers own shopping habits (at a variety of online vendors providing True Fit services) and comparing purchase and return histories to known sizing and style attributes, True fit can help their direct clients improve both profit and consumer loyalty, as well as instigating a protention brick and mortar relationship (when possible) which is likely to be 5 times more profitable.
True Fit Analysis - What True Fit Gets Out of It
Who are your taget users? The target users in-house are the sales, marketing and loyalty teams.
What are they key goals?
The key goal of internal partners is to build a clearing house of vendor specific and vendor agnostic data which they can use to strengthen relationships with corporate partners to help create new data products such as targeting fit to underserved markets, product design suggestions, sales data about consumer actions and connecting online to brick and mortar.
- How can you help them accomplish those goals?
By building depth in data stores, True Fits recommender team can create leverage for their sales and support team to build more durable client relationships and work more closely on custom products.
The Recommender Itself
This recommender system for true fit appears to look at (based on their descriptions and behaviors of the system) online and offline purchases and returns. The system evlauates the sizes, styles, fits, fabrics, economics and demographics of each purchase to try to establish what fits us physically, emotionally, visually and financially and then build recommendations based on these purchases.
The recommendations take into consideration global purchase and return habits weighted by your specific behaviors and answers to questions in your own True Fit account. Over time, the more purchases you make from True Fit vendors, the more specific your recommendations become.
One of the biggest advantages True Fits model has over other online fashion recommenders is that their algorithm has actual physicological data about human shapes and sizes, and data about the dimensions, fabrics, cuts and give of shoes and clothes on their sites (including the compettitors of the current site you are shopping) which it has created factors for that they can evaluate relative to what you look at, buy and return to all the sites you frequent and even some brick and mortar stores.
Because much of their data is proprietary and it is not fully clear how user data is integrated into fit data to make recommendations, it is not possible to create a simulated model of their system, but it would very likely start with a giant matrix of physical dimesions as attributes as well as style related infomration, colors, patterns, fabrics and such and then models about your clothing buying habits would be comapred to this matrix to make recommendations.
Where data is thin, you would likely be compared to customers who had made similar or the same purchases and been offered the products which they also purchased and kept. Using a model like this you could leverage the habits of customers across brands and vendors improving the quality of recommendations as customers purchase and return more.
Over time, your clusters of customers could become rich enough to not only recommend clothing to customers, but potential products to vendors where there are gaps in their lines!
Citations
True Fit? | Genome. Consumer profiles, Transaction data, Fit & Style Data Aggregated, Normalized, Mapped. (n.d.). Retrieved November 8, 2017, from https://www.truefit.com/Products/Genome
True Fit Introduces Genome™ for Footwear and Apparel. (2015, October 28). Retrieved November 8, 2017, from http://www.businesswire.com/news/home/20151028006121/en/True-Fit-Introduces-Genome%E2%84%A2-Footwear-Apparel