8 Ways AI Can Boost Retail Sales During the COVID Holiday
With Cyber Monday upon us, the 2020 holiday shopping season is now in full swing. Thanks to the COVID-19 pandemic, online shopping will see major growth, but retailers of all types should be ready for major changes. The big question–and the big opportunity–is how companies can leverage data and AI technology to survive and then thrive.
The global retail market is worth a mind-boggling $26 trillion per year. Most of that commerce will continue to be done in person, as much of the world is not set up for UPS deliveries. For online retailers, business is growing fast, and under COVID-19, they’re growing even faster.
Consider the case of a little-known online bookseller, Amazon.com. As its sales ballooned by around 40% in the second and third quarters, thanks to the novel coronavirus, it hired about 250,000 new employees to handle the business surge, leaving it more than 1 million employees and on a pace to soon unseat Walmart as the largest employer in the country.
The pandemic has accelerated many companies’ ecommerce and digital transformation plans, in many cases by several years. That makes it even more important for companies to optimize their online presence to maximize sales, which means being intelligent and opportunistic in how one approaches ecommerce.
Here are eight ways big data and AI tech can help you accelerate your ecommerce initiative:
1. Sweet Product Recommendations
Giving people what they want is a great way to attract new customers and to retain existing ones. Ideally, one knows something about their customers, or can make intelligent inferences to surface products or services that people may would be inclined to buy. In each case, data and machine learning play critical roles in putting the right offers in front of the right customers at the right time.
It’s not just a theoretical improvement, as studies have shown that serving good product recommendations can have a major impact on sales. According to a study by Invesp, 45% of customers say they’re likely to shop on a site that offers personalized product recommendations, while 56% are more likely to return to such a site.
However, there’s quite a bit of room for improvement. According to a 2018 study by Bazaarvoice, only about one quarter of shoppers are treated to a personalized homepage when shopping online. That’s actually good news for companies looking to broaden their reach, as it shows them what low-hanging fruit is available to boost sales.
2. A More Intelligent Search
Using a search engine is perhaps the most natural thing that people do on the Internet, and it remains a primary tool for ecommerce shoppers everywhere. Considering the ubiquity of the technology, one would think that it has been perfected by now. You would be wrong.
“Search is far from being solved,” Ciro Greco, the director of AI for intelligent search firm Coveo, recently told Datanami. “It’s the hardest thing we do. It’s the hardest thing everybody does.”
Surfacing useful search results is as much an art as a science. It’s not easy and requires much more than the installation of a search engine that uses natural language processing (NLP) technology. Today, there are a number of software firms angling to deliver the best offering in this category, defined as Cognitive Search or Insight Engines, that help with the full spectrum of search optimization, including data integration and user behavior analysis.
3. One to One Marketing
When Don Peppers and Martha Roger’s published “The One to One Future” back in 1994, the concept of marketing directly to an individual was more theory than fact. Fast forward 26 years, and the advent of advanced analytics and AI have put micro-segmentation and hyper-personalization within the reach of marketers everywhere.
With one to one marketing, companies are take a more targeted approach in delivering a personalized experience than they would with personalized product recommendations or search engines.
Data, such as page views and clickstream behavior, forms the bedrock foundation of one to one marketing, according to an article by Dynamic Yield’s Shana Pilewski. As data is processed, commonalities emerge that correspond with broad segments. As the data is further refined, a picture emerges of an individual’s preferences, which can inform real-time action.
4. Finding the Perfect Price
There are a lot of variables that influence a consumer’s decision to buy something: product availability, seasonality, size, color, etc. But numerous studies point to price being the number one factor determining whether a purchase will be made.
Pricing is a practice that has traditionally been done by analysts after studying reams of data. But companies are now increasingly turning to machine learning-based methods to accelerate the analysis of all that data and create an optimized price, sometimes several times a day. This helps to keep the price in that sweet spot where it’s not so high that it turns off potential buyers or cannibalizes other products, but it’s high enough to deliver a good profit.
There are various approaches to using AI for price optimization, as Gabriel Smith, the Chief Evangelist and Vice President of Innovation at Pricefx, wrote about in Datanami last month. Not every company has enough data to power ever approach, so finding out which approach works for you is likely a process of trial and error.
5. Inventory Optimization
Earlier this year, at the beginning of the COVID-19 pandemic, American retailers ran out of inventory of crucial items like face coverings, bath tissue, and hand sanitizer. In some product categories, the supply chains have still not recovered, eight months later. Nobody could predict the emergence of the coronavirus and its impact on retailers, but the current episode shines a bright light on the need for better inventory optimization and planning in the consumer goods supply chain, regardless.
Retailers and distributors that adopted machine learning-based approaches for supply chain planning did better than their colleagues who continued to rely on human analysts equipped with Excel. Nobody has gone completely unscathed, since the pandemic was not predictable, but having a pre-built planning model to work from and to be able to tweak is providing to be a crucial tool for brick-and-mortar companies to survive COVID-19 long enough to be able to recoup some lost sales during the annual end-of-year spending extravaganza.
With a model in place, the data can lead to better decisions. “Incorporating external data modules like social media data (Twitter, Facebook), macro-economic indicators, market performance data (stocks, earnings, etc.) to the forecasting model, in addition to the past samples of the inventory data seasonality changes, best determines the product demand pattern,” writes Zensar engineer Kalvoju Vivek in a recent blog post.
5. Going Omnichannel with Fulfillment
Retailers have never been so challenged as they are today, particularly with the coronavirus making a late-season assault on North American and European homes and businesses. In many parts of the United States, county health departments have instituted occupancy limits of 50% or less of regular limits.
That is pushing many people who would normally shop in person to shop online instead. But retailers should consider other options for getting goods into the hands of its customers, including contact-less transactions and curbside pickups.
These omnichannel fulfillment patterns were in place before the pandemic, but they’ve accelerated greatly under COVID-19, according to Matthew Shay, president and CEO of the National Retail Federation.
“Innovations are taking place in a matter of just months that would normally take years, in areas like acceleration of e-commerce offerings, blending of digital and in-store experiences, curbside pickup and quicker delivery options, and contactless delivery and payments,” he says.
7. Fight the Fraudsters
Fraud is always a concern around the holidays. But thanks to the COVID-19 pandemic and the resulting shift to online everything, fraud levels have increased 60% this year, according to Feedzai’s Quarterly Financial Crime Report. “COVID has created a big disruption in the banking, payments, and e-commerce sectors with multiple impacts all over the world,” says Jaime Ferreira, the senior director of global data science at Feedzai.
According to Feedzai, a projected 30% increase in card-not-present transactions sets the stage for fraudsters to order goods from cards that have been compromised. But there are plenty of other avenues for fraudsters to work their nefarious magic, including card skimming, lost and stolen cards, phishing scams, account takeovers, and application fraud. All told, fraudsters are projected to increase their take-home by about 5.5%.
Luckily, card issuers and merchants alike are armed with machine learning, which gives them the capability to analyze huge volumes of transaction, identify the instances of attempted fraud, and automate the response to it. AI spending in cybersecurity is already a large market, and it’s projected to grow at a healthy 23% yearly clip, becoming a $38 billion market by 2026, according to Markets and Markets.
8. Bring On the Bots
Chatbots that can automatically respond to common customer requests are one of the fastest growing segments of big data and AI. Thanks to improvements in NLP and natural language generation, chatbots can understand written and spoken words at levels never seen before.
They can save companies millions of dollars per year lot of money by eliminating the need for human customer service reps (or allowing the humans to move into areas that provide a higher value-add, as the rumors say). You may not personally like chatbots, but they’re the future of customer service for a variety of industries, including retail and ecommerce.
However, business should be careful with their chatbots. According to Forrester Research, 63% of customers will leave a company after just one poor experience, and almost two-thirds will no longer wait more than 2 minutes for assistance.