×


GET IN TOUCH WITH US
Boston | London | Los Angeles | Berlin | Nice | Mumbai

Corporate HQ: 617.848.3740
General Inquiries: 877.FIT.TRUE

HQ: Boston
60 State Street
12th Floor
Boston, MA 02109
View Map

Europe HQ: London
1st Floor, Golden Cross House
8 Duncannon Street
London, WC2N 4JF

Contact Us

We'd love to hear from you.

×

Schedule a Live Demo

 

True Fit CAO, Chris Moore on The Impact Podcast: Aggregating Massive Data Sets

September 24, 2018

the-georgian-podcast.jpgA few months ago, True Fit’s CEO and Founding President, William Adler, was featured on Georgian Partners’ Podcast, The Impact. Adler discussed the ways in which True Fit is working to solve some of the biggest issues in the fashion industry through machine learning. He also discussed the ways in which the fashion industry is changing, and how True Fit maintains the largest customer data collective.  

True Fit’s Chief Analytics Officer, Chris Moore, was featured on The Impact as well. He discussed the importance of normalizing data, some of True Fit’s best practices for managing data, overcoming the challenges of working with thin data, and much more. Here are some of the key takeaways from Moore’s discussion:
 

The Importance of
Data Normalization

Moore is adept to the importance of data use and engagement. Previously, he worked in a variety of different industries, proving the importance of data across varying fields. While data collection and analysis is highly important, Moore stresses the importance for businesses to effectively normalize their data.

“Data normalization is the whole name of the game,” says Moore. “You can’t really scale a business like this unless you can take the data from all the crazy places you get it, and normalize it into one system where you can really work with it and scale that business without having to redo all the analysis for every client individually.”

In fashion, Moore explains how each individual product True Fit analyzes has 140-150 data points on it. Some items, such as dresses, can carry even more data points than the average due to varying silhouettes, colors, fabrics etc. With the help of data normalization, True Fit is able to effectively leverage all of this specific data into one database, called the Fashion Genome, for customer analysis and accurate product recommendations.

“You can’t really scale a business like this unless you can take the data from all the crazy places you get it, and normalize it into one system where you can really work with it and scale that business without having to redo all the analysis for every client individually.”

Training Data to Personalize

Screen-Shot-2018-09-26-at-9-48-47-AM-(1).pngGathering data from the retail manufacturers begins the process helping customers find apparel and footwear that they will love and ultimately keep. Normalizing the data helps set up the machine-learning software to provide recommendations based on the preferences of each unique customer.


“It’s really about trying to produce what we call features, that are relevant to the problems you’re trying to solve and which you can extract from every possible example of the clothing you’re working with. That enables you to convert clothing to a set of numbers really about all the features that garment has. And if you’re good at your feature engineering you can help a machine learning algorithm get at the underlying truth of what’s going on,” said Moore.

Proper Product Recommendations

Through proper machine-learning techniques, True Fit works to provide accurate apparel and footwear size recommendations. However, an equally important part of the larger goal is introducing customers to new products that they may like in terms of fit and style, based on previous purchases.

“What I’m trying to do is line up the properties of things you like and know with things you don’t know yet and help introduce you to that,” says Moore.  

True Fit also analyzes customers over time, in order to consistently provide accurate recommendations based on preferences, which have the likelihood to change over time. The more recent purchases made by customers take slightly higher importance, due to this being the closest indicator of the customer’s current preferences, and measurements.

Looking to learn more about the importance of proper data aggregation?
Click here to listen to the podcast in its entirety or listen to the podcast below