Stitch Fix – Driving Retail Profits with Data Analytics and Mass Customization

Stitch Fix has innovated upon the retail business and is attempting to escape from fast fashion operating problems by using data analytics to offer personal service to customers and real-time feedback to designers.


Stitch Fix Logo
[Photo: Stitch Fix (5)]
Stitch Fix is a fashion retailer which has gained notoriety for its unique business model and its deep resonance with consumers. By branding itself as a personal styling service rather than a retailer, it has unlocked a well of data to drive mass customization allowing it to build customer loyalty, forecast demand, and ultimately generate high margins.

How it Works – Consumer Side

For a $20 fee, Stitch Fix sends its users a box containing 5 items of clothing, a “fix,” either on a pre-arranged date or on a monthly basis. If the user buys any one item in the fix, the $20 fee is applied to the cost of the item and the user receives a 25% discount for purchasing all 5 items in the fix. Shipping is free both for delivery and returns. (1) Stitch Fix items have an average price point of $55 and the company expects its users to keep two items out each fix. Stylists are paid $15/hour and are allowed to complete 4 fixes/hour leading to high gross margins which cover substantial overhead inventory costs.

How it Works – Company Side

Stitch Fix Style Sample
A Sample Style Profile [Photo: Business Insider (2)]
Beneath this business model, the interplay between a data analytics program with uncanny accuracy, a clever distribution network, and a few shrewd but critical business decisions drives the company’s success.

Although Stitch Fix emphasizes the personal connection between stylist and customer, the stylists are actually choosing from a small subset of products recommended by its proprietary data analytics engine. New customers fill out an extensive set of questions stating their size at several common stores and indicating preferences for certain “styles” by numerically ranking photos of different sets of items, eliminating potential miscommunications and creating a more approachable aura.

Data like these as well as Pinterest Boards, Instagram posts, and Facebook likes are aggregated with information about the users previous likes and dislikes (for repeat users) to generate a set of potential items. One over 1000 stylists selects the final 5 items that go into the fix and prepares a personal note suggesting why the items were chosen and addressing any special requests made by the users. The program is sophisticated enough that it only shows the stylist items that are in the user’s size and price range and which will be in stock during the time when the fix is scheduled. (3) Over time, Stitch Fix builds up a library of likes and dislikes so that users become more and more satisfied with each passing month, adding to the value created and customer loyalty.

Inventory management poses one of the most interesting operational challenges of this business model. Stitch Fix expects consumers to keep two out of the five items they are sent, so on average an item will be sent out two-three times before being purchased. Customers have three days with the merchandise, so to minimize inventory holding costs, Stitch Fix has established a network of distribution centers which allow it to log merchandise back into the system more quickly and decrease the garment turnaround time. (4) The 25% discount if all five items encourages marginal purchases, eliminating the cost of shipping and turning around only one item in all but the most extreme cases since it is often cheaper to purchase all 5 items than only 4.

Stitch Fix also strikes an interesting balance when purchasing inventory. On one hand, consumer value would be the greatest with a huge library of individual pieces but it would be incredibly expensive to curate that broad collection. On the other hand, there would be tremendous cost efficiencies in a model which stocked a very deep collection of only a few styles. At present, Stitch Fix has struck a balance by carrying over 200 brands, with an average per item price point of $55.

Shrewdly though, they have eliminated one of the biggest risks shouldered by more conventional clothing retailers through a deeper understanding of the customer. (2)  The average customer is a woman between 25-40 and over 50% of customers have at least one child. This consumer is not so much concerned with being at the forefront of fashion as with not being too far outside of the mainstream. She does not want to be the first to have a hip new item, but rather to hop easily onto a rolling bandwagon. Stitch Fix does not have to worry about trends. It purchases a mix of new items and items which did not sell well at the first retailer (similar to TJ Maxx or Marshall’s) and then fills in obvious gaps with items from its own 6 fashion lines. The company also scrupulously avoids small demographics, declining to offer plus sizes (for which it has attracted criticism) and avoiding extremes at either end of the price point scale. (3)

How it Works – Partner Side

On the retailer side, Stitch Fix is able to leverage its data to provide value to designers and wholesalers. Traditionally, retailers might have relied on direct feedback from only a few customers to determine why a given item did not sell and wholesalers would have no information. Stitch Fix offers specific reasons why women are not buying specific pieces of clothing, in real time, and can use this information to negotiate lower prices. (3)


(1) Simone, Ahuja.  “What Stitch Fix Figured Out About Mass Customization.” Harvard Business Review Online Article. Posted:  May 26, 2015 Retrieved: December 8, 2015. URL:

(2) Peterson, Hayley. “This Hot Fashion Startup Eliminates the Hardest Part of Shopping.” Business Insider. Posted: March 12, 2015. Retreived: December 8, 2017. URL:

(3) Petersen, Anne Helen. “Can Silicon Valley Fix Women’s Fashion?.” Buzzfeed News Features. Posted June 4, 2015 Retrieved December 8, 2015. URL:

(4) Colson, Eric. “How Does Stitch Fix Handle Its Inventory Management?” Answer to Posted Question on Quora.Com, Response by Chief Algorithms Officer or Quora. Written: March 17, 2014, Retreived: December 8, 2015. URL:



SpaceX and the Operating Frontier


The Milky Way: The evolving National Dairy Development Board of India

Student comments on Stitch Fix – Driving Retail Profits with Data Analytics and Mass Customization

  1. Thanks Rachel! I’ve been meaning to try out stitch-fix, but I couldn’t find all the information I would need in one place. 🙂

    I enjoyed learning about how StitchFix provides data to retailers. Retailers clearly want the data very badly, which can be seen whenever they offer prizes for filling out reviews of their products online. However, aside from an award that I’m not going to win, I have little incentive (other than guilt – because I use reviews) to write any reviews of my own. With StitchFix, the act of writing a customer review creates value for the customer, for StitchFix, and for the retailer. It’s a really smart concept!

    As a member of their key demographic, (correct age range, no kids, most often just trying to fit in), my main concern would be realizing the clothing being sold to me didn’t sell well initially. When I’m in the mode of just trying to fit in, I’d start to question my “own unique style” if I were liking all the clothing that no one else wanted. I feel, though, that people don’t talk about StitchFix in the same way that they talk about TJ Maxx, which has some great finds and some really awful ones, and which I certainly think of as a disorganized discount place. You’re definitely right that they seem to be a personal styling service rather than a retailer. The thought that someone picked out these items with me in mind, even if it’s actually mostly an algorithm, is the factor that would allow me to trust the service.

Leave a comment