Insights

Retail at crossroads: Navigating AI and changing consumer demands

Peter Charness, VP Retail Strategy - UST

On August 11, 1994, a transaction occurred that would change the face of commerce forever. Dan Kohn, through NetMarket, sold Ten Summoner's Tales, a CD by Sting, to Phil Brandenberger of Philadelphia using a credit card over the Internet. This transaction is credited as being the first secure online commerce transaction. Within five years, Amazon was selling books, eBay was selling various items, Dell was custom-building and delivering computers, and PayPal was innovating financial transactions. This new model laid the foundation for online B2C commerce.

In the 1990s, US retail was dominated by department stores and mall-based specialty stores. Retailers were embracing technology, albeit at a slower pace than other industries, and building hardened, high-scale, low-cost systems, processes, and supply chains. Just-in-time (JIT) inventory management was prevalent, and customer satisfaction was the watchword. Brand owners had to be extraordinarily cooperative with retailers to get their products on the shelves. Traveling the world as a retail buyer was considered a prestigious career, and retailers could easily find people to fill these roles.

Fast forward to today, and the landscape has dramatically shifted. Of the top 40 department stores in business in the late 1980s, only seven still exist under their original form and ownership. In 2020, influenced by COVID-19, about 30% of all retail business was transacted over the web, with an even higher percentage at least influenced or researched electronically. Hardened supply chains with JIT inventory management crumbled like bowling pins. Retailers realized that supply chain processes need to be less rigid and more agile, the exact opposite of their previous construct. Relentless customer focus remains the mantra, but it's an era where customer satisfaction (I want what I want, and I want it delivered by noon) and transactional economics are at odds. Retail systems are "old enough to vote," and the rush to omnichannel/unified commerce capability further challenges dated and disjointed architecture. Today, retailers must be nice to brand owners to get the best products on their shelves. While retail buying is still a good job, experienced people are retiring, and replacements are scarce.

AI/ML, and particularly Generative AI, have made significant strides in recent years. Technological breakthroughs enabled model training with parallel data loading, which, combined with better specialized hardware, provided the basis for Chat GPT to garner over 100 million downloads in just 60 days. In August 2023, it saw over 50 million unique visitors a day.

In 1995, anyone who predicted that less than 20 years later, delivery trucks would be making daily trips to households that had researched and ordered products on their phones, would have been considered crazy.

Today, numerous predictions exist for the future use of AI in business. Although there are already some clues, anyone who thinks they fully understand the business and societal impacts that will take place is likely both wrong and underestimating the changes. Unlike the internet's role as the fundamental enabler of a new world of retail, which arguably took place over 15 to 20 years, a good prediction for AI might be that whatever transpires will happen much, much faster.

Retail today is at a crossroads. At the risk of repeating the mistake of the 1899 US Patent Office quote, "Everything that can be invented has been invented," it seems logical that the next big gain for retail may, in fact, need to be a lot of little gains.

DIVIDER

Small gains, big impact

While not true for all retailers, most can further optimize supply chains, but is there one more big advance or system out there that will fundamentally reduce time and cost? Forecast-based replenishment solutions can always improve, but the biggest reductions in inventory happened when the first replenishment solutions were implemented. Price and promotion management solutions have already yielded significant margin gains. The next levels of capability will provide more benefits, but likely one customer, one store, one offer at a time, with no wholesale single improvement.

New product time-to-market: integrated computer-assisted design/lowest cost/best global production provided the big bang, which will be refined from here, with smaller improvements to be gained.

DIVIDER

The power of micro decisions

As the above examples indicate, the next wins in retail may come a penny at a time, improving costs and operations one store, one customer, one order, one delivery, one search or ad, one truckload, or one home delivery at a time. Retailers need to harvest millions of wins that, in the aggregate, will yield large financial results. The next "big thing" will be a lot of little ones. How will the same staff manage even more?

DIVIDER

The loss of tribal knowledge

Great retail practitioners are becoming scarce. Those with many years of experience are set in their ways and time starved. Worse, many great practitioners are taking their knowledge home with them as they retire at a high rate. We need to equip the next hire with the tools to be productive, intelligent and informed on how to do their jobs. The tools need to "teach retail" to fewer workers who will have to manage even more with fewer team members, but with better tech.

DIVIDER

Escalating customer demands

Every time a retailer raises the bar on seamless service, customers expect the same everywhere. Failing to deliver or innovate is failing your customer. All customer touchpoints need refinement. Retail systems that deal with most of these needs remain difficult to learn and hard to adapt or change to support the next functional need. Large-scale modernization projects or solution replacement projects are often characterized by overly lengthy implementation, cost overruns, and lack of adoption.

DIVIDER

The future is now

This is the reality for most retailers today. Retail in 2030 will be an AI-infused experience that will make it easier for customers to say yes and for fewer retail workers to get better, smarter, even faster. Retailers face the difficult challenge of adopting technologies that are not quite ready for prime time but essential for success in the near future. Like the 1995 prediction, the future is murky and impossible to predict with high accuracy. However, if you're not already at least experimenting with some of these new capabilities, you're late to the starting line. Chances are 2030 will arrive at your business even sooner than six years from now.

Learn how to be future-ready in retail at https://www.ust.com/en/industries/retail-and-cpg

DIVIDER

Resource

https://www.ust.com/en/industries/retail-and-cpg

https://www.ust.com/en/alpha-ai

https://www.ust.com/en/our-partners