Retail has always operated with thinner margins than many other industries. It made up for lower prices and the high operational costs of brick and mortar with volume. Before the Internet, when someone needed school supplies, groceries or a new television, they drove to a local shopping center and purchased them. The ebb and flow of retail almost directly matched the ebb and flow of the overall economy.
That is no longer the case. For more than a decade, traditional retail has been in decline. Between 2010 and 2017, an estimated 12,000 brick and mortar stores have closed shop, with major brands like Toys R Us and Sears – once mainstays in American shopping centers – declaring bankruptcy and closing most or all of their stores. But despite the cries of a “retail apocalypse”, people still buy things. While eCommerce is a huge part of the retail die off, it represented 14.3% of total retail sales through the end of 2018. Expected to reach 22% in 2023, eCommerce continues to carve out space for itself in the retail landscape, but savvy, data-driven retailers who have thus far weathered the shifting winds and have a plan can still thrive.
One of the greatest assets in business is data, and retailers have an immense store of this “digital gold” with which to operate. Those that invest in predictive analytics and build a product strategy driven by observation can overcome the margin busting impact of online shopping and thrive in the 2020s and beyond.
Data is all around us. It always has been. But technology now makes it possible to tap into the wealth of information we have at our fingertips to make smarter, more calculating decisions. Netflix built an entire business model on predictive analytics – evaluating what people watch, when they watch it, and for how long they watch. They then develop programs based on those insights. Major League Baseball’s executives have completely changed how they do business thanks to sabermetrics – a series of statistics that dig deeper into player performance than the sport ever had before.
At its most basic, predictive analytics is the process of reviewing data points about an individual’s behavior and determining what they are most likely to buy as a result. It takes into account things like:
Exactly when do people buy things and what impacts those results? Is it strictly about the calendar, or does the weather have a similar impact? It’s easier than ever to break down this data with machine learning algorithms.
With the average retailer recording 20% of their sales during the holidays, it’s more important than ever to target the right people at the exact right time. How early do people buy? What do they buy? Jewelers, for example, see an even larger spike in holiday sales and they tend to come slightly earlier than average holiday spending.
With the right data, marketers can build smarter offers and distribute them to the appropriate channels, inventory can be established and managed months in advance to avoid shortages of popular products, and revenue can be forecast more accurately to start preparations for the next fiscal year in advance.
We’ve seen the basic ideas behind demographic targeting for decades. Advertising on TV typically matches the stereotype of the audience expected to watch. Ads for expensive luxury cars, watches, and razors accompany baseball games, while toys and video games accompany children’s programming.
But we can dig deeper now with predictive advertising. Who is the ideal prospect for a new BMW? Where do they spend their time? What content do they read? Data tells us exactly where to run ads and what to accompany those ads with.
Inventory is one of the biggest challenges faced by retailers. The annual additional cost of holding too much inventory is between 25% and 32%. It can destroy margins and, especially for products that have a narrow window of relevance before they become obsolete, the cost can be even higher.
Predictive analytics allows marketers and retail managers to determine what products need to be in stock and when they should be in stock. If one product goes out of stock, data can help determine what the next best option will be as well, creating a stocking plan that reduces the potential for excess inventory.
This goes beyond ordering for individual stores and can help restructure the entire supply chain through discriminant analysis and computer vision.
Consumers increasingly expect a highly personalized experience. In part, this is due to massive retailers like Amazon, which drives more than a third of its sales through its proprietary recommendation algorithm.
The year it launched, Amazon increased its revenue by 29%. Upsells aren’t new in retail, but they have the potential to be smarter and more personal than ever before, delivered directly to consumer smartphones, email inboxes, and even live on the website as they shop.
Amazon has long experimented with dynamic pricing. You likely see this whenever you login – prices are constantly changing to find the sweet spot. Moreover, their recommendation engine leverages these changes to send emails about recent price drops, evaluating how much of a price drop will trigger purchases. For high volume products, this kind of dynamism can be a powerful tool in eCommerce.
Predictive analytics allows marketers, supply chain managers, and even store managers to make smarter data-driven decisions that improve organizational performance.
As the old rule of thumb goes, 80% of sales come from just 20% of SKUs. If analytics can help narrow in on that 20% of SKUs, reduce inventory on the remaining 80% and build a smarter supply chain and marketing machine to promote those SKUs to the right people, the efficiency of retail can be greatly enhanced. That’s where retailers will ultimately find their weapon against eCommerce – a smarter bespoke experience for those who still prefer offline shopping.
In 2019, Artificial Intelligence and Machine Learning are moving beyond trendy experiments and becoming core strategic elements in businesses of all sizes. To learn more about the state of AI in 2019 and how it is influencing industries like healthcare, banking, media, journalism, and marketing, download our recent report, AI and Machine Learning Trends in 2019. Learn where investments are happening, and the way technology is transforming businesses.
It’s this evolving technology landscape that is the focal point of the upcoming DATAx conference on November 6-7, 2019 in New York City. Bringing together technology and business leaders from the world’s top companies, DATAx is about identifying and implementing real-world solutions to common challenges through AI and machine learning. You’ll hear from data science leaders at Bank of America, The Associated Press, Airbnb, Nasdaq, WW (Weight Watchers), and many more on what their teams and companies are doing to leverage AI and innovate.