by Pranav Tyagi
This is a pivotal time for retailers. Many were already facing pressures from the growth in ecommerce prior to the onset of the pandemic, and after months of stay at home orders and forced closures, these stresses have only increased. According to recent data from IBM’s U.S. Retail Index, COVID-19 has expedited the shift away from physical stores to digital shopping, moving the timeline up by nearly five years.
As physical stores have reopened, it has become apparent that consumer preferences have changed for good. Brick-and-mortar retailers looking to survive and thrive in this dynamic landscape need to develop a new playbook, incorporating robust omnichannel strategies, as well as incentives to pull consumers into their physical storefronts. Those brands that take an active approach in retooling their strategies will fare much better than those that try to resist the shift.
How Machine Learning Can Help
Before retailers can make any necessary changes to their strategies, they need real-time, updated data to make informed decisions. While relying on last year’s data to help illuminate the path forward was previously effective, the pre-COVID-19 consumer looks far different than today’s. Operators need to reexamine how they are collecting data, what technologies they are using to analyze it, and how they are aligning those insights with location strategies moving forward.
Advanced technologies that include artificial intelligence and machine learning provide retailers with more comprehensive analytics around key data elements, such as shopping preferences and behaviors, changing demographics and local traffic. This allows executives to make more informed decisions specific to store locations and maximize sales and profitability at a market-level.
Machine learning provides the ability to sift through old and new data sets to find the best combination of models to interpret the current environment. By using these techniques, retailers can process large volumes of information and discern patterns that are beyond the capabilities of human analysis, especially when the underlying data is changing frequently.
These techniques are tailor-made to address the rapidly changing impact of closed competition or shifting consumer patterns, and provide operators with forward-looking location-specific sales forecasts that reflect the current environment. Retailers can employ a myriad of useful data points relating to key store locations and consumer buying patterns that are vital to understanding which locations are positioned to perform well, which should be closed, and which need new strategies or relocation.
Having the right game plan can make all the difference, as we learned from Target’s success. Target was at the forefront of developing omnichannel strategies with a comprehensive rollout of Buy Online, Pick Up In-Store (BOPIS) fulfillment centers across a wide number of store locations. To anyone who has experienced Target’s BOPIS fulfillment centers, it is clear they are designed efficiently with an unimpeded flow of consumer traffic. BOPIS shoppers have the option to pick up their goods without interacting with other patrons, or if preferred, they can just as easily head into the physical store to continue shopping. This sort of design does not happen by accident, but it is the marriage of meticulous planning and excellent data. The planning seems to be paying off as Target announced plans to expand BOPIS offerings to include grocery items this year.
The Value of Speed
It is not just the comprehensive nature of machine learning models that make this technology so well suited to the new retail environment. Instead, it is the speed at which information is processed and analyzed. If having the right information is the most important factor when it comes to navigating the post-COVID-19 landscape, accessing it in real-time is a close second. The ability to instantly pivot or address pain points is vital when looking to retain or grow your customer base in a competitive and disruptive environment.
Dunkin’, the popular coffee franchise, is a great example of a retailer that used intelligent information to quickly pivot during stay at home advisories in March. The coffee giant quickly instituted plans to offer curbside pickup and promoted its delivery capabilities via popular apps such as Grubhub, UberEats, and DoorDash to ensure customers could continue to purchase products while remaining socially distanced. Dunkin’ continues to find ways to streamline consumer purchasing, having introduced an initial pilot program for its checkout-free stores in September.
The Power of Flexibility
Another important advantage of machine learning is the ability to easily recalibrate. Attempting to alter and recreate models manually is laborious and error-prone. This could prove costly for retailers relying on the validity of old models, as recent events in October show the volatility of 2020 has no sign of slowing. With the possibility of market realities once again shifting, it is vital that analysts are able to alter models to include shifting consumer sentiment or unforeseen variables. Whichever way the pendulum swings, retailers need to be certain they can keep their models current so they are able to best align their strategies. Making a sub-optimal location decision is a 5 to 10 year mistake, so getting it wrong is not an option.
Physical stores will continue to be a vital part of successful retail strategies, but operators need to be tapped into shifting consumer sentiment. The second wave of the pandemic and what may prove to be a long, challenging winter will mean continued fluctuation in consumer confidence and buying habits. The retailers best positioned to succeed in this environment will be those that have their fingers on the pulse of consumer needs and are able to address those dynamic pain points quickly and efficiently.
Pranav Tyagi is the Founder, President and CEO of Tango.