The Future of Gen AI in Retail: Balancing Human Factors With Revenue Growth

If you buy into the exploding hype around generative AI, you see this flavor of artificial intelligence as the best thing to happen to computing since cloud storage. But you could be getting the wrong impression.

As businesses and industries continue to evaluate the pros and cons of ChatGPT, generative AI, and other artificial intelligence species, some adopters praise its time-saving and innovative benefits. Others are hesitant to trust the new technology. Either way, where gen AI is headed is an ongoing conversation.

Talkdesk in January released a report warning that continuing bias and inaccurate data have seeped into retail experiences, already integrating AI, and are impacting consumer attitudes towards the new technology. This belief comes as shoppers chronicle AI-powered interactions gone amok and worry about how businesses may use facial recognition, customer data, and general unethical AI use cases.

Shoppers, already dissatisfied with their customer experiences, are ready to leave behind any brand not practicing responsible AI use, according to the Talkdesk Bias & Ethical AI in Retail Survey. At the same time, corporate officials rave about how targeted their AI results are and backslap one another in excitement that everything is peachy keen.

This mixed sentiment may not bode well for an accelerated expansion of gen AI this year, as some proponents predict. The report reveals some shocking and negative attitude changes in how consumers feel about interfacing with AI, Shannon Flanagan, VP and GM of retail and consumer goods at Talkdesk, told the E-Commerce Times.

Her company provides a cloud contact center platform for AI-powered customer service.

“I have definitely seen an attitude shift. There is some shocking information about how gen AI is being used for product recommendations that people are not using. And then there are high expectations shoppers have on data security and transparency not being met,” she offered.

AI and Gen AI – What’s the Difference?

Artificial intelligence has been quietly deployed with limited capabilities for nearly a decade. Its use cases in recent years have gradually improved thanks to advancements in machine learning (ML) and combination with robotic process automation (RPA).

The release of ChatGPT last year marked a significant breakthrough that enhanced automation for repetitive, rules-based activities requiring minimal human oversight. This advancement has broadened AI’s capabilities to encompass a more comprehensive array of functions.

All AI programs are not the same species. Traditional artificial intelligence focuses on analysis and classification. Generative AI, or gen AI, is a subset of artificial general intelligence technology that uses complex algorithms and neural networks to simulate human creativity and produce new content from models that can include text, images, sounds, animation, 3D models, and other types of data.

Gen AI captures nuances in language and generates output based on the patterns on which it was trained. Its models can remember previous interactions, resulting in more coherent and relevant conversation experiences for users.

However, gen AI cannot make decisions involving many complex factors. At least not yet. It excels at making data-based suggestions but is inept at including and handling the most critical human factor.

Putting AI Into Productive Practice Can Fall Short

Research shows that disconnects exist in how businesses can safely and accurately integrate gen AI skills into their business cycles and avoid unintended consequences. In retail and call center circles, consumers are not unanimous about how AI is affecting their customer experiences (CX).

Flanagan has seen a definite shift in user attitudes as the Talkdesk platform integrated gen AI capabilities. Not all the changes reflected in the company’s numerous surveys favored AI.

“Some of our pre-holiday AI surveys talked about how shoppers are feeling about AI versus retailers. A large part of them aren’t doing it,” she told the E-Commerce Times.

Big brands like Walmart are legitimately using gen AI. However, according to Flanagan, a broad spectrum of her company’s clients do not know how to utilize it.

“Product description copy is kind of a no-brainer. In some places, the use case in customer service is a no-brainer. But there’s still a lot of hesitancy,” she said.

Consumer Sentiment Toward AI

The recent Talkdesk report unveils startling findings on AI’s use in product recommendations, revealing a majority of surveyed individuals are not utilizing them. Additionally, consumers highlighted unexpected demands for data security and transparency.

Flanagan emphasizes the urgent need for a strategic overhaul to engage customers effectively, pointing out the now evident use cases.

Still, there are issues about using AI that must be resolved, she cautioned. That fix should be easy to accomplish, especially in uses that serve as agent assistants rather than customer-facing integration.

“Now the reality is for doing it in anything that’s customer facing. It should be seamless to the customer, but that’s a little bit more risky than I think some of the back office uses like a marketing ops, and then obviously like the agent assistant in our self-service world,” Flanagan explained.

Examples from the Talkdesk report about how shoppers use AI show that:

• 79% of shoppers refrain from purchasing AI-powered product recommendations because they are not customized to their interests;
• 71% have never purchased a recommended product because it makes them feel like a brand is monitoring them;
• Only 28% of those surveyed believe that retailers are being safe and smart with their data.

“As it relates to AI, this is a ton of mistrust,” she observed. “What needs to be done this year is a little bit of a pause and saying, what’s our strategy?”

Another Study Shows AI Paying Off

Yet another prominent AI report gives a much different view. According to a new study from MessageGears, 99% of marketers say that using AI has impacted their ability to understand customer preferences and behavior.

A key takeaway from the survey of enterprise marketers in companies with 500 or more employees is that the vast majority already use AI in their marketing, which is paying off. For today’s marketers, the big goal is making real connections with customers. Doing so strengthens brand recognition and builds trust.

Bottom line: Surveyed enterprise leaders said AI has been particularly useful in boosting customer engagement.

“AI algorithms are like the secret sauce, letting marketers dive deep into customer data,” Will Devlin, VP of marketing at MessageGears, told the E-Commerce Times.

“Then, armed with the inside scoop on preferences, behaviors, and demographics, marketers can fine-tune messages on the fly. With real-time tweaks to content, timing, and more, AI-powered campaigns ensure that the connection between a brand and its audience is spot-on and meaningful.”

Conflicting Results Skew AI Assessment

Only 53% of marketing pros surveyed by MessageGears said they were very successful at connecting with customers. This statistic leaves significant room for improvement.

Another 53% want to use the tech to identify more accurately who will most likely make a purchase. Half would like AI to help them pinpoint the most effective channels to reach customers.

Fifty-eight percent of marketers in the MessageGears survey use AI in targeted advertising campaigns. Almost half (49%) use the technology for personalized email marketing, customer support and service, and customized product recommendations.

Further, 97% of enterprise marketing experts using AI said they successfully delivered personalized content and recommendations, while 39% said the experience was exceptional, and 99% said AI is making a big difference in figuring out customer preferences and behavior.

A critical use of gen AI from a marketing perspective in 2024 will be to help fix the customer engagement problem. Customer engagement is all about providing value and communicating that value to your customers in a way that makes them feel connected and appreciated, offered Devlin.

“Your customers should be excited about what they’re receiving from you. Messaging should be timely and relevant and be delivered on the channels that are most important to your customers. Businesses already know this, but it’s often a manual guessing game to make it happen,” he told the E-Commerce Times.

Devlin added that marketers should anticipate increased utilization of predictive AI and modeling to determine the most effective communication strategies with customers, eliminating the need for guesswork. Marketers can then pair these predictive AI insights with generative AI to further refine and personalize the message.

AI Enterprise Growth

ChatGPT’s first anniversary marks the remarkable aspect of the ascent of generative AI, marveled Priya Vijayarajendran, president of technology and CTO at gen AI software developer ASAPP. She stressed that its democratization and ability to unite talents from various corners of the technology landscape allows the best and brightest to harness their skills to collaborate to “get it right.”

“Moving forward, responsible usage of data and investment in AI privacy and assurance are essential so we can unlock the potential of generative Al for enterprise innovation. This innovation must continue. There is no slowing down now,” she told the E-Commerce Times.

Generative AI will continue to deliver incremental innovation across GPUs, LLMs, and compute frameworks; she said of expected progress this year. Data will dominate as the most significant differentiator, applying LLMs in a hybrid domain focus to achieve accuracy, time to value, and scale.

“These vectors coming together will be the key to unlock exponential value [of Gen AI] for enterprises,” Vijayarajendran concluded.