Times are tough for H&M. As a result of powerful trends, namely the rise of online shopping, the world’s second largest clothing retailer has seen its fortunes wane.
As noted by the Wall Street Journal‘s Saabira Chaudhuri, the company has endured 10 straight quarters of poor same-store sales, has seen its stock price cut in more than half and has engaged in price reductions to move billions of dollars in unsold merchandise.
Part of H&M’s problem is that it has historically stocked its stores with similar merchandise, regardless of store location or size. That has fast become a liability in the age of ecommerce, as consumers no longer have to buy whatever local brick-and-mortar retailers decide to carry. Instead, shoppers have no shortage of options for finding the products they really want online and having them delivered to their door.
To address the impact this is having on its business, H&M has started taking its vast trove of data, which covers everything from sales, store visits and returns, and running it through algorithms that the company hopes will help it figure out what merchandise should be carried at each store. It is also looking at external data sources, such as blog posts, in an effort to identify hot new trends months in advance.
As the Wall Street Journal’s Chaudhuri points out, H&M’s fine-tuned approach, while sensible in theory, is still unproven.
There are also lots of questions about AI’s abilities and limitations. For example, H&M’s algorithms had suggested that the retailer promote reindeer sweaters in January because the algorithm didn’t realize that sales of such sweaters had only increased in the run-up to Christmas for reasons that are obvious to a human.
Such issues aside, there are signs that adding AI to the mix could be fruitful. For example, based on what it discovered through data, H&M slashed the number of items it carried at its store in Stockholm’s Östermalm neighborhood by 40% and focused on products for women.
H&M says that the store’s sales have improved significantly, but did not provide the Wall Street Journal with any numbers.
A tale of two brands
Taking H&M’s performance claims at face value, it would appear that big data and AI could be a boon for the company. But the big question is whether or not it will stem the tide or be the start of a meaningful comeback.
The answer to that question isn’t so clear.
H&M’s biggest fast fashion competitor, and the largest clothing retailer in the world, Zara, also faces many of the same competitive threats as H&M, but it’s thriving.
There are a variety of reasons for this. Some of them are structural. For example, a majority of Zara’s factories are located in proximity to Europe; less than a third of H&M’s are.
H&M also has nearly twice as many stores as Zara – 4,288 compared to 2,127. While omnichannel retailers are increasingly finding ways to use their physical locations to their advantage, that seems to be too many stores, especially when one considers than in the US, H&M’s retail locations are heavily weighted to malls.
But H&M’s issues might not only be structural. Pamela Danziger, president of marketing consulting firm Unity Marketing, argues that Zara is beating out H&M because it focuses on experience, not product, and places more emphasis on value than price.
She also suggests that Zara puts the customer first, which helps it turn customers into evangelizers, whereas H&M puts its brand first and tries pushing its customers through promotions.
Finally, she believes that Zara “is way ahead” with its omnichannel strategy.
The holistic fix
Can H&M’s bet on big data and AI can help it right its ship? Perhaps. But in analyzing why it has struggled while its biggest competitor has thrived, it seems clear that data and algorithms alone won’t be enough.
While these can certainly help provide insights and drive optimizations, the structure of a business and the strategies it adopts ultimately provide the foundations on which data and algorithms will operate, so getting these right is not an option.