Zara Leverages Data Analytics to Understand Consumer Tastes

Zara is using data analytics to guide its design and manufacturing process, making it a leader in the fashion market.

Zara is a key player in the fast fashion retail business. It got to this position thanks to its approach towards data analytics. Zara is huge, In FY 2014 it sold $19.7bn worth of merchandise, slightly behind H&M’s $20.2bn and ahead of giants such as Uniqlo ($16.6bn) and GAP ($16.4bn) ( And sales are constantly growing, in the first half of 2015 Zara revenues were up 17% compared to H12014. Behind these impressive numbers is a revolution in the world of fashion, which Zara is spearheading. Zara and its fast fashion competitors are leveraging data analytics to change the approach towards supply chain management and inventory.

Traditionally, retailers would do their best to estimate demand for the different SKUs, mostly based on industry experts’ opinions. Next, retailers would manufacture in China or another low-cost market, stock stores, and, well, that’s it. From then on models that sold well would quickly be out of stock, and models that didn’t sell as well would have large inventories and later on be discounted and sold with thinner margins. Early on in the season retailers had some room for re-stocking, but because of the long cycle time from order to distribution, the ability to adapt to demand was quite limited.

Zara solved this problem with an adaptive, data-driven supply chain management. Zara’s process starts in a similar way to the traditional retailers – with an initial order. The difference is that instead of ordering the bulk of the quantity for the season, Zara only orders a small amount of merchandise. Once the merchandise hits the stores, Zara collects sales data and analyzes each SKU’s sales against supply. Zara does even more, it analyzes performance of features of different SKUs. For example, they might identify that pants with patches sell better than pants without patches, or that certain colors or fits move faster than others. Zara then uses these insights to guide their following orders. They will design and manufacture models that have the most popular features to satisfy demand.

The key to making this process happen is short-cycle, small-batch manufacturing. Instead of in the far east, Zara manufactures about half of its merchandise in company-owned facilities in Spain and Portugal (, reducing the production cycle from a few months to a few weeks. While this might be a more expensive production process, Zara still succeeds in maintaining profitability – Inditex’s (Zara’s parent company) gross margin was 56.9% in Q22015, compared to GAP’s 37.4% ( Zara is able to maintain these high margins and capture significant value by (1) reducing the amount of inventory and the cost associated with holding excess inventory, and (2) selling less items on discount thanks to its flexible production process.

With this data-driven approach Zara is able to create more value for its customers. Zara gives customers the models they want, when they want them. While in the traditional world the most popular models are quickly out-of-stock and only the duds are available, in the world of fast fashion supply is highly adaptive and caters to the evolving taste of the consumers. Customers know that items they buy at Zara are new and trendy. As mentioned above, Zara is able to drive value capture by selling more full-priced items and running less discounts. Customers know that batches are small and if they don’t buy the item they like now it will soon be gone (replaced by other trendy, popular items, but still gone). Thus, they will be less likely to wait for end-of-season sales and will be more willing to pay full price.

While Zara was a pioneer in leveraging data to guide production, many companies have been shifting towards the fast fashion model. For example, Uniqlo has recently entered the US market and plans to further expand globally. Uniqlo leverages the fast fashion model but it also has a strong portfolio of basic items that provides stability and is less privy to effect of trends (e.g. the Uniqlo sweater collection). Zara’s ability to remain a strong player in the market depends on many factors (e.g. brands strength, cost structure, marketing) but it also heavily depends on the quality of its insights, which helps drive up margins. As more and more companies adopt this model, Zara must further develop its data-analytics skills and demand-prediction abilities to remain ahead of its competition.


Palantir – Masters of Big Data


The NYC MTA: “Keeping Up” in the Age of Information

Student comments on Zara Leverages Data Analytics to Understand Consumer Tastes

  1. Thanks for the great post, Noam! I completely agree with you that Zara has grown into this huge apparel brand by leveraging data. The company has been extremely adaptive to customer tastes and this has helped them deliver the growth that they have so far. What makes me bullish on Zara going forward in a world of e-commerce is this very strength in data analytics. While traditional brick and mortal stores which have not focused till recently on digital are struggling to make a transition into the the digital space, Zara given its traditional strength in digital and data analytics will be able to counter the threat from e-commerce players. A strength which has helped them grow in the face of incumbents will not come to their rescue again when the disruptors are attacking them!

  2. Really interesting post! I wonder if part of Zara’s success is due to the fact that the small initial orders drive heightened demand, as consumers realize that each piece will sell out quickly so they should buy it as soon as they can. E-commerce retailers like Reformation do this well by starting with a small amount of inventory for each piece and then as soon as it sells out, they leave the piece on their website but write that it is “sold out” and ask interested customers to sign up for the waiting list. This increases consumer desire and creates a cycle of high demand and low supply, which keeps people interested. It will be very interesting to see how retailers like Zara continue to use data in the future!

  3. Great post! This post made me Google around a bit to better understand Zara (and other retailers’) demand prediction models. For the statisticians (or just plain nerds) out there, this link does a good job of explaining the basic premise of Zara’s regression models.

    I wonder if the brand’s e-commerce channel can serve as another input to the demand prediction model. By tracking which products, colours, and designs are performing better than the others in which regions, or running A/B experiments, a retailer can successfully connect the offline with online, assuming, of course, that the preferences and behavior of the online shopper are the same as that of a retail store shopper.

  4. I truly enjoyed reading this post, Noam!
    Zara is ahead of its peers in many aspects. You put it into a very nice perspective! I am wondering how Zara should approach the sales. To maximize profits, you obviously want to sell majority of goods for full price (and creating a “it may sell out feeling” is a great push factor). At the same time, today’s retail customers are so used to sales. When a new sale wave starts at Zara, the stores are crammed all day long. If Zara succeeds in making the production process very short and accurate, where will the discounted merchandise come from? Will Zara shorten / abandon the long weeks of low prices that happen several times a year? What is your opinion?

  5. Loved reading about Zara — definitely a great example of just how important data is in developing a competitive advantage in mass retail. Going forward, though, I feel like this is something that all mass retail brands will need to employ — seems like it’s already happening as you mentioned with Uniqlo and others. Given that fashion tastes of consumers change so rapidly, I wonder whether having been first to market in using data will grant Zara a long lasting advantage — aside of course from having developed internal data analysis capabilities. Google, for example, benefits from having collected data far longer than anyone else, making their algorithms impossible to compete with. In Zara’s case, their algorithms are probably more or less created with a new season’s designs. Will be interesting to see how they stay ahead of the competition.

  6. Very interesting. Zara’s vertical integration and speed of response is difficult for other competitors to replicate. Furthermore, the company understands that consumer behaviours are too fickle and leave the decision making to consumers. It’s interesting to watch a fashion company that executes so well on its operations and demand forecasting.

  7. I think Zara’s integration of data with supply management could provide a example of how traditional brick-and-mortar stores could respond to the e-commerce disruption. However, is Zara successful because of the special attributes of the fast-fashion industry? Could other traditional retail businesses, like Wal-Mart, that could do what Zara is doing and be competitive enough to make its brick-and-mortar stores profitable despite the sales-loss that is being taken away by e-Commerce?

Leave a comment