destinacigdem/Depositphotos.com
June 11, 2025
Is AI Fatigue Setting In as Retailers Struggle To Scale or Incorporate Tools?
Artificial intelligence (AI) has captured the imagination of many business sectors, including retail. During this year’s National Retail Federation Big Show held in New York City, AI tools — and topics ranging from gen AI to agentic AI, and their use cases — were the unambiguous throughline dominating so many discussions.
Now, as Fortune pointed out, “AI fatigue” could be setting in at the corporate level.
Citing data from S&P Global Market Intelligence, Fortune writer Sage Lazzaro indicated that the share of companies that had nixed the bulk of their AI initiatives had surged from 17% in 2024 to 42% so far this year. Perhaps even more striking, data suggested that the average company had scrapped nearly half (46%) of its AI proofs of concept instead of executing them.
What’s more, Lagazzo pointed to a Quantum Workplace study which painted a somewhat counterintuitive picture of AI incorporation into enterprise workflow: Employees who rated themselves as frequent AI users reported higher levels of work burnout (45%) than those who infrequently (38%) or even never (35%) leverage AI for work-related tasking.
Lagazzo quoted Erik Brown, the AI and emerging tech lead at consulting firm West Monroe, as to why this trend appears to be growing in stature in 2025.
“Anytime [that] a market, and everyone around you, is beating you over the head with a message on a trending technology, it’s human nature — you just get sick of hearing about it,” Brown said.
“I think it’s so easy with any new technology, especially one that’s getting the attention of AI, to just lead with the tech first. That’s where I think a lot of this fatigue and initial failures are coming from,” Brown added.
TechCrunch provided a parallel data point when referencing recent transaction data from fintech company Ramp in a June 9 report. In that report, the outlet noted that Ramp’s AI index — “which estimates the U.S. business adoption rate of AI products by drawing on Ramp’s card and bill pay data” — saw a flattening rate of adoption as of May 2025 following 10 consecutive months of growth.
In May, 49% of large businesses had incorporated AI in some fashion (as well as 44% of medium-sized companies, and 37% of smaller businesses). And although TechCrunch AI editor Kyle Wiggers did point out the incomplete portrait being painted by Ramp’s AI index — which draws from corporate spend data sourced from ~30,000 companies), he did note that there was one general truth in evidence.
“But it’s certainly true that businesses are beginning to realize that there’s a limit to what today’s AI can do,” Wiggers concluded.
Retailers Struggle To Scale AI, Despite Widespread Experimentation
And according to a May 30 blog post from Valtech, citing information gleaned from a retail roundtable with Google Cloud, retailers are also exhibiting a halting of progress in terms of scaling AI tools.
The cornerstone data point provided by the consulting agency: While nearly 85% of retailers are currently experimenting with AI, only a fraction of that number (a sizable minority of 15%-20%) have “scaled it in a meaningful way.”
The three primary causes of failure-to-launch when it comes to artificial intelligence tools, as outlined by Valtech? A lack of clear ownership, meaning that AI pilot projects often originate in tech or data teams but are missing an experienced and capable business lead who can take ownership and drive value from the tool; a lack of repeatability, which sees so many successful pilots display early promise but then fizzle out due to a lack of designed scalability; and unclear or fuzzy metrics, because without useful feedback loops based on a constant flow of data or valuable input, there’s very little incentive to continue deployment.
Matt Hildon, retail portfolio director at Valtech, laid the majority of the blame for failed AI scaling in retail at the foot of individual businesses who weren’t prepared to execute.
“The tech is ready. The business often isn’t,” Hildon said, somewhat arguing against Wiggers’ position.
Discussion Questions
Is AI fatigue, when speaking of retailer deployment or scaling of AI tools, setting in? If so, what signs of this are you personally observing?
Are the limitations of the current crop of AI models becoming more evident, or are companies (and their personnel teams) simply not leveraging the tools to maximum effect? Can both be true? Do you identify more with Wiggers’ position, or with Hildon’s?
Are consumers, themselves, exhibiting signs of fatigue or skepticism surrounding agentic AI when deployed by retailers, in your opinion?
Poll
Loading …BrainTrust
Mark Ryski
Founder, CEO & Author, HeadCount Corporation
Patricia Vekich Waldron
Contributing Editor, RetailWire; Founder and CEO, Vision First
Shep Hyken
Chief Amazement Officer, Shepard Presentations, LLC
Recent Discussions
More Discussions
Would Recording In-Store Associates Help as a Training Tool?
Ethosphere is banking on AI-driven recordings of in-store associate practices as the next big thing in…Some Retailers Are Trimming Holiday Sales Discounts: Is This a Dangerous Play?
Some retailers are pulling back from deep discounts this holiday season. Is this a dangerous play?Has Kroger Discovered the Limits of Automated Warehouses?
Kroger is winding down its relationship with Ocado, citing cost overruns and a reorganization of its…
The truth is that AI has been overhyped, as most retail technologies are. It’s not that it’s irrelevant or that it won’t have an impact – it isn’t, and it will. The problem is twofold. First, AI does not replace getting the retail basics right – which so many retailers still fail to do. AI can’t rescue any retailer from that. Second, the promise of AI currently exceeds its practical applications. That means retailers need to be selective in how and where they deploy it. There also needs to be a focus on making it work because what businesses need in this environment is tangible results.
I can’t speak for retailers, but I’ll speak for RetailWire readers (at least one of them): fatigue is definitely here!
Seconded.
In just the past month, Hudson’s Bay Company, the oldest retailer – in fact oldest company. period – in North America was liquidated, at the hands of American owners (the same group of geniuses [mis] running Saks). To a large extent this ends department stores in Canada. One might think that all the boxes this checks would have generated at least one thread from RW…instead we got 389 on AI.
Yes.
catchall phrases (ai, machine learning, large language, metaverse) are trite very quickly.
Why?… as not discussed in tactical way with measurable result.
Shiny objects don’t hold a crumbling world together.
Purpose (action & outcome) is necessary.
Tech is pushed into the world when not ready for primetime.
The ability of AI to improve outcomes and reduce costs is real. But AI is also being over-hyped. Following the Gartner Hype Cycle we’re coming out of the “Peak of inflated expectations” stage and moving into the “trough of disillusionment.” The “Slope of enlightenment” is ahead. That’s when crazy talk becomes outcome-focused implementations. Not everyone is on the same cycle timeline.
The AI euphoria is wearing off, and reality is setting in. Retailers can’t simply AI their way to success. Identifying use-cases and experimenting is one thing, but implementing at scale successfully is an entirely different and more challenging exercise. AI is everywhere, but the barrier is retailer execution. Constraints on time, money, expertise, clean data, all play a role in stunting value realization and meaningful outcomes. Retailers need to take a breath. Focus on the critical few areas that they see the greatest potential benefit and work on executing these areas. Be patient and practical.
“Retailers can’t simply AI their way to success.”
The inherent issues remain, and are complex.
You can process data 24/7 and barrage your highest value customers with messages… but if they don’t like the product in store… they don’t.
The speed of data… vs… speed of supply chains/contracts/delivery/merchandising
AI. AI. AI. AI. Aw you tired yet?
While AI certainly doesn’t solve all woes, especially those of strong execution in a retail store, I’m not fatigued by it at all. We are just at the beginning of the potential of AI, and as someone who has spent hundreds of hours learning and testing it in a variety of ways, I’m excited!
The fatigue is from generalities. Not the relevant areas of application (powered by ai): eg. fraud detection, medical diagnostics, and customer service automation
I have never seen a technology as overhyped as AI. It’s a cool thing, but I don’t see it having the kind of impact in retail as it will in other industries.
it’s just settling into reality.
Consumers – especially young consumers – are willing to try and use AI. However, it has to be easy and frictionless. Anything that requires extra effort of any kind will taint the experience. For the retailers that get it right, they set the bar for others. So, find the retailer that best represents the experience you want to create. Finally, and this is my new mantra about AI… Even though AI can do something, that doesn’t mean it should.
An industry colleague recently posted this on LinkedIN:
“So we had ….
… the Internet revolution
… the mobile revolution
… the cloud revolution
… then the SaaS revolution
and now the AI revolution !!
At every single one of these foundational technology changes consulting firms, CIO, analysts said the same thing.
1- the technology is more advanced than the ability of companies to use it
2- technology is not the limitation – the issue is people, org structure and incentives
3- it’s all about the business process – you need to reimagine that. “
Fatigue settling in? Yes, but the problem remains that we keep talking about the same hard truths and dressing them up as revelation. Let’s stop admiring the problem. We need to mix these truths into playbooks that move the needle. Not just the conversation. In my experience, real innovation has come from turning “friction points” into “design levers”.
Less hang-wringing, more systems-thinking. As such, both “AI model limitations” and “teams not using to fullest potential” can be true. What is the business problem we are solving for? How do we measure it? Do we have the data? Do we have the right resources and infrastructure to invest towards them?
If not, “hit next” and go to next initiative. You are not ready, and to your final question – the fatigue, will turn criticism and show itself in the voice of the customer/consumer/shopper/patient/etc.
Yes! We continually rediscover the same fundamental truths about technology adoption, often presenting them as new revelations.
You are right on, Alex. You have been quiet, and I understand why. Keep forging that new pathway forward for both customers and partners. As you say, it will take a village to rid us of this fatigue.
I certainly am weary of hearing the hype – and so are retailers judging from the reactions at recent meetings and events. Like other new technologies, most have experimented with it, and many are using AI in targeted functions of their organizations. Over time the hype will balance out and it will be used by organizations that can identify where it adds value and that can execute.
“many are using AI in targeted functions of their organizations”
Focus on the purpose.
Focus on the business/brand/consumer problem.
The discussion needs to be tactical. (Leave the letters AI to be the power supply, like electricity running but not the end goal).
If you think it is overhyped you haven’t seen the amazing things it can and is doing. It’s much more than a bot or automation. I say the hype is matched by the expanded ideas of retailers as they use their new tools.
Retailers tend to ask, “How can we use AI?” when they should ask, “What customer friction can we eliminate, and what repetitive employee tasks can we automate?” There are many similarities to prior transformational technologies that have gone through the “trough of disillusionment” when reality collides with hype, like e-commerce adoption in the late 1990s and ERP implementations in the early 2000s.
Most retailers haven’t mastered basic data hygiene, customer segmentation, or operational excellence that AI requires to function effectively. Yet, current AI models do have limitations and struggle with context, causation, and the messy realities of retail operations.
Consumer fatigue is setting in with the poor deployment of AI. They don’t complain about Amazon’s recommendation engine or Netflix’s content curation. Still, there are enough chatbots that can’t understand context, recommendation engines that suggest irrelevant products with personalization that is more creepy than helpful. Fatigue can set in anytime when hype and reality collide. Still, the heart of the matter is whether retail organizations will mature their strategies and adoption processes to match AI’s advancing capability.
“Most retailers haven’t mastered basic data hygiene,”
As contextual use is still in infancy!, maybe the core use should be in the longstanding data gathering/storage/protection area. As well as the hygiene (relevancy, recency, consent levels).
This is typical and driven by the industry that hypes trends to drive engagement. The focus should always be on the problems we solve. I feel that NRF this year started showing maturity, moving away from tech terms to focus on the problems it solves. The fatigue is setting in as customers maturity increases as they learn more about what is real and available as a solution, versus aspirational and experimental. This is why we need to always evaluate solutions based on how Accurately, Repeatedly, at Scale, and Speed they solve the problem.
I agree with the majority here, AI should be used as an awesome assistant and to create efficiency, not replace the basics or humans. AI should not be a distraction from the real work and conversations that need to be done or had.
This is not fatigue so much as “early adoption blues” which happens pretty much anytime a technology has been overhyped and overpromoted prior to proper stress testing and formulation of a business plan that adds value to the retailer and their customers.
Further much of what consumers “feel” regarding AI deployment is the continued, exceedingly stupid and tedious voice response nonsense you get anytime you need help.
Slow down!
Yep. The industry will soon start applying realistic and very narrow use cases, the data will show promise in revelations, interpretive dashboards will be built, and good decisions will come.
It is not the least but surprising to me to learn that half of corporate AI initiatives so far have died on the vine. This is exactly what we should expect from any revolutionary tech innovation. This is not failure. It is proof that companies are investigating and taking measured risks to incorporate new technique. Learning and failing fast is is a necessary feature of tech adoption.
To the point about “lack of clear ownership”: Let’s be clear, retailers are not pivoting to become AI entities. They are incorporating AI-enhanced tools to improve their core business — which last I checked, was still procuring and selling merchandise to shoppers.
AI tweaks are coming in every functional area of the business, as tech developers add capabilities designed to improve accuracy and speed decision-making. Ultimately it is up to the functional experts in each area of a retail company to define which AI tools are hype and which truly help.
Since this is still a new realm wise retailers might benefit, for a time, by establishing temporary “AI centers of excellence” to help top managers make informed decisions about this.
AI fatigue is less about the technology and more about how it’s being used. When pilots lack clear goals, ownership, or a plan to scale, teams can lose momentum. AI holds massive potential for retail, but the implementation needs to be practical, measurable, and aligned with business goals from day one.
The hype is out there, but retailers/grocers need to give AI time to be properly applied, tested in learning stages, with good data sets for starters. It’s not an overnight endeavor with ROI out of the gate.
If retail execs are fatigued, then they’ve likely either scoped AI projects with grandiose outcomes, are partnering with a vendor who doesn’t/won’t understand retail/grocery intricacies, or aren’t applying the right metrics to narrowly defined data sets and outcomes. Or all the above.
I’m bullish, have seen it tested real time in Grocery retail. AI application must start small and needs a caretaker within who will own it, communicate clearly at the top levels of leadership.
While AI fatigue may be setting in among some retail stakeholders, retail giants are doubling down on this technology. Amazon, Walmart, Target, Shopify, Starbucks, Nike, Sephora and McDonald’s are among the leaders investing in AI-driven efficiencies.
We are experiencing significant changes with AI adoption, and the concept of AI fatigue is emerging. Commentators suggest that like the Internet, organizations that embrace AI will be the most successful. Effective AI deployment should shield users from having to choose between models. Though overwhelming, retailers like UK’s Pets at Home have gained industry recognition and new opportunities through AI use. Anecdotally speaking, it feels to me that very often, ambitious AI adoption is driven by strong leadership or a visionary individual recognizing its potential.
The idea of “AI fatigue” really resonates – there’s so much hype, but not every solution delivers real value. In retail, I’ve found the most effective tools are the ones that solve specific, day-to-day problems. For example, using a solid versandsoftware https://shipstage.com/ like Shipstage has made a huge difference in managing our shipping logistics. It’s not flashy AI, but it’s smart, practical tech that actually helps the business run smoother. That’s the kind of innovation retailers need more of.
AI fatigue has been studied for a while now, just suppressed to get that VC gravy train moola pouring in, check out ai-fatigue.com/ there are even some case studies posted on amazon: https://www.amazon.com/AI-Fatigue-Artificial-Intelligence-Employee-ebook/dp/B0D2BQV1DC/