How retail brands can survive the attack of the algorithm

In a world where big data is king, brands need purpose to stand apart

By Published on .

Credit: Illustration by Patty Alvarez

The combination of algorithms and retail is irresistible. By matching supply and demand as closely as possible, they learn about our preferences and shopping behaviors, predict what we'd like, develop tailored recommendations—and increase the likelihood of a transaction. That's why there's an entire new industry emerging around shopping aggregators and wardrobe-curation services.

By turning products, services and content into commodities, algorithms, by default, are the opposite of brands. Some investors have even concluded that algorithms will replace the role of brands in conveying trust and guiding us through our choices. That's why the modern test of brand strength is its resistance to the algorithm.

The true challenge for retail, then, is not how to accelerate algorithm-induced product commoditization, but how to circumvent it.

Otherwise, retailers will find themselves on the same path of self-destruction as publishers, which put their faith in the hands of Facebook and Google algorithms. The ongoing digital media carnage is brought on precisely by the belief that on the internet, anyone can be a content creator, and the algorithms will do the rest. When the race turned into search results and clicks, everyone joined the business of clickbait and covering the same things. The competition for indistinguishable offerings became stiff, and many couldn't survive it.

And just as "there were too many of us doing the same job" in publishing, as Caroline McCarthy writes in her Spectator USA article, "How Digital Media Killed Itself," in retail there are now too many newcomers offering the same things. And much like Google and Facebook, Amazon won't help them differentiate.

Amazon is for the products of the world, but "Amazon is not a home for brands," says Melanie Travis, the founder of swimwear brand Andie. "They basically want to reduce us from a brand to a product." Once that happens, it will make these products directly compete with Amazon's private-label business.

Algorithms also don't exactly level the retail playing field. Similarly to how Google and Facebook favor popular content, retail algorithms favor the most affordable, frequently purchased and frequently reviewed items, often preventing smaller brands from being discovered.

Amazon has also never overcome its fake review problem, and it's not alone. Last fall, cosmetics brand Sunday Riley drew attention for encouraging its employees to post fake product reviews on Sephora. Meanwhile, a recent New York magazine article, "How Much of the Internet Is Fake?" noted that less than 60 percent of web traffic is human; click farms, bots and fraudulent traffic account for the rest.

In this context, retailers increasingly realize that the best way to pass the Turing test is to be human. Companies that have a point of view and stand for something beyond what they sell are more likely to influence purchasing decisions, attract and retain customers and create a long-term competitive advantage.

According to Accenture Strategy's Global Consumer Pulse, 63 percent of global consumers gravitate toward buying products and services from companies that have a purpose and share their personal values and beliefs.

Products are a way to express identity, belong to a community and convey one's passion and beliefs. Patagonia and some of its customers share values of social responsibility. In comparison, Stitch Fix's "save time, look great" pitch doesn't seem to give it the same lift, at least when looking at the brand's engagement rate on its Instagram account. (StitchFix has an engagement rate of less than 4 percent; Patagonia's is above 5 percent.)

Despite the hype, retail algorithms are mostly incremental improvements on the traditional retail model where the core user activity is product purchase. Brands today are in the race of offering more. The next wave of retail disruption is happening not in the domain of targeting and efficiency, but in the domain of purpose.

Most Popular
In this article: