Is AI adoption moving too fast?
Photo: Getty Images/FluxFactory

Is AI adoption moving too fast?

According to KPMG’s “Thriving in an AI World” study, 53 percent of business leaders in retail said COVID-19 increased their pace of AI adoption, yet 49 percent believe adoption is moving faster than it should in their industry.

The broader survey of 950 business and/or IT decision makers found similar sentiments for other industries, including industrial manufacturing, 55 percent; and tech, 49 percent. The concerns were traced to debates surrounding the ethics, governance and regulation of AI.

Among retail respondents, 78 percent said it is difficult to stay on top of the constantly evolving AI landscape — a sentiment higher than leaders in other categories. Cybersecurity breaches (47 percent) and possible AI bias (45 percent) were found to be the top two greatest potential risks of AI adoption. Eighty-seven percent believe the government has a role to play in regulating AI technology.

Matt Kramer, national consumer and retail sector leader at KPMG, said in a statement, “The concern about the speed of adoption raises a caution flag for retailers, reminding them to ensure proper process and controls along with change management and effective training are put in motion to address the AI adoption risks.”

Among the encouraging responses from retailers:

  • Eighty-one percent said AI is moderately to fully functional in their enterprise, up 29 percentage points from a year ago. 
  • Seventy-eight percent said their company has an AI code of ethics, up 34 percentage points;
  • Ninety percent indicated that their employees are prepared and have the skills for AI adoption, up 47 percentage points.

A recent survey from Rackspace Technology of 1,870 organizations across industries showed only 17 percent of companies have mature AI/machine learning capabilities, but other adoption challenges were cited.

The top five barriers to actionable insights were data quality issues, cited by 31 percent; followed by lacking the capabilities to effectively manage our data, 30 percent; data being stored in too many different systems, also 30 percent; inability to process data quickly enough to act on it in time, 29 percent; and difficulty collating, structuring and integrating data in a meaningful way, 28 percent.

BrainTrust

"While the rapid pace of AI adoption seems overwhelming, the pandemic revealed the urgent need for new efficiencies. "

Lisa Goller

B2B Content Strategist


"One thing is clear, the retailers who are adopting AI and using best practices today will be tomorrow’s winners."

Matthew Pavich

Sr. Director Retail Innovation at Revionics, an Aptos Company


"Are we talking about true AI or AI in name? Many tech companies commonly call automation AI today in the race to attract both talent and funding."

Karen Wong

Co-Founder & CEO, TakuLabs Ltd.


Discussion Questions

DISCUSSION QUESTIONS: Has the pandemic spurred the adoption of artificial intelligence technology at too fast a pace? Are cybersecurity breaches, AI bias, data quality and access or other factors the biggest hurdles to overcome?

Poll

30 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
David Naumann
Active Member
3 years ago

AI, done right, can yield much quicker and better decision making for companies. It was surprising to see that 81 percent of retailers have AI moderately to fully functional in their enterprise, up 29 percentage points from a year ago. I think the two biggest challenges for adoption of AI are the lack of trust in letting AI make decisions that were historically done by humans and the lack of quality data. From a data perspective, with data silos and poor quality data, retailers need to make sure they have clean data and a single, consolidated enterprise-wide data repository to help avoid “garbage in, garbage out.”

Zel Bianco
Zel Bianco
Active Member
3 years ago

Some may think that the industry is going too fast, but in the world of e-commerce there is no way you can collect, process and generate insights from all that data without AI/ML. Yes there are areas that are messy and need to be improved and optimized but learning along the way is better than not being in the game. It’s like the saying: fail fast, and often, but if you constantly improve the process you will be in a better position to win. The data is messy and AI can help to find the gaps that would take people way too long and be too expensive to scale.

Mark Ryski
Noble Member
3 years ago

I seriously doubt that the pace of “meaningful” AI adoption is too fast or that the pandemic has spurred adoption. Most retailers I encounter are still dealing with day-to-day operational issues resulting from the pandemic. And while I have no doubt that “AI” gets bandied about and that there’s plenty of experimentation and dabbling, this is hardly meaningful. The applications for AI are many, and it will continue to be applied and evolve, but in my experience working with retailers this is still more of an aspiration.

Cathy Hotka
Trusted Member
Reply to  Mark Ryski
3 years ago

Agreed. This isn’t the first survey we’ve seen that touts retailers’ prowess in this area, and I suspect that the retailers who are using AI to great effect don’t want to brag about it for fear of tipping off the competition.

Bob Phibbs
Trusted Member
3 years ago

Why do all stories about AI treat it like a new flavor of Oreo rather than a program? AI is only as good as what it is taught to execute, measure, and connect. It is bespoke, not mass market and getting clarity on what people think AI is and does is the first step.

David Weinand
Active Member
3 years ago

Boy, that’s not a phrase you hear very often: “adopting technology at too fast a pace”! The data quality issues are real and to fully take advantage of AI capabilities, this has to be addressed. Also, the AI term has been added to just about every tech solution out there so it’s hard to know what is really AI and what is marketing. Ultimately, the benefits of machine learning/AI are adding real value to companies so I don’t think it is being adopted too quickly – it’s just that it’s not likely adding full value until data issues are addressed.

Ken Morris
Trusted Member
3 years ago

I don’t believe AI is moving too fast but quite the opposite. What we are moving too slowly on is addressing the silos that are impeding AI, the use of clean and real-time data, and the taking of prescriptive action based on that data. We need to fund the infrastructure that allows us to thrive by leveraging our most important asset. Systems that are a day late in synchronization just won’t allow you to complete tomorrow. We must normalize our data and stop creating islands of information that don’t talk. This takes investment but the prize is worth it.

Lee Peterson
Member
3 years ago

OK, I’m going to play the Digital Immigrant here and ask for an official definition of AI. I mean, a calculator is AI. My kick-ass laptop is AI. Heck, Google is AI! If the real definition is simply “advanced technology,” then I’m afraid any idea of timing is WAY too late! We are tumbling down that road so fast you’d better ask Siri quick what NFTs are before you go broke!

Dr. Stephen Needel
Active Member
3 years ago

I’m betting many of these respondents have no clue about what AI actually is and that most are thinking of it in terms of automation rather than generating something new to think about. I say that because there is no way 90 percent have their employees ready to deal with AI adoption. So is it moving too fast? Yes if it’s AI, no if it’s automation.

Gene Detroyer
Noble Member
Reply to  Dr. Stephen Needel
3 years ago

I love your comment. You are exactly correct. Say “AI” and people get the willies. Say “automation” and they say “go, go, go!”

Venky Ramesh
Reply to  Dr. Stephen Needel
3 years ago

Great point. Let’s understand first what AI means – I would call it a collective term for capabilities shown by learning systems that humans perceive as intelligent, that they would normally expect from only a genius. This can be achieved through machine vision and sensing, natural language processing, the ability to crunch data gathered from all senses to derive meaning and to predict, etc. Robotic automation is the most menial form of AI and even that hasn’t seen enough adoption.

Mark Ryski
Noble Member
Reply to  Dr. Stephen Needel
3 years ago

Couldn’t agree more Steve.

Shep Hyken
Active Member
3 years ago

AI is here and it isn’t going away. It creates great advantages for both retailers and consumers. The concern for privacy was already here. Retailers must do the right thing by the customer to earn and keep their trust.

Suresh Chaganti
Suresh Chaganti
Member
3 years ago

This probably indicates treating AI as a technology initiative as opposed to a transformational one. When the implementation is not cross-functional, the outcomes will be disappointing.

Lot of people don’t realize, but AI is only as smart as the humans that configure the AI algorithms. And those humans need domain expertise, business acumen, and executive sponsorship to follow through the hypothesize/model/test/refine cycle. The results will not be pretty in the beginning.

The survey probably indicates the feeling that the full cycle of implementation methodology is not used, even though they may not be able to articulate it.

Venky Ramesh
3 years ago

in the hope of seeing some AI at work, I recently visited Walmart’s AI lab called Intelligent Retail Lab (IRL) which is hosted in a neighborhood Walmart store in Levittown, NY, and was surprised to see the store fitted with thousands of cameras and other equipment but no sign of them being used. Speaking to some store employees, I learned a lot of theoretical plans of how AI will be used in the store (shelf scan, smart replenishment, etc.) but no one was able to recall being assisted by AI. Perhaps someone from Walmart can clarify. On the other hand, Amazon is making big progress with Amazon Go and Amazon Fresh and using some aspects of AI on their e-commerce site (smart returns for example). For the number of tech capabilities that exist today, the progress just isn’t enough- too slow. It’s definitely not for the dearth of use cases. I am guessing the cost is a big factor when it comes to scaling AI.

Jeff Sward
Noble Member
3 years ago

I suspect that the 49 percent that are worried about AI moving too fast are more accurately worried about their company moving too slowly and will soon experience a competitive disadvantage as a result.

Karen Wong
Member
3 years ago

Are we talking about true AI or AI in name? Many tech companies commonly call automation AI today in the race to attract both talent and funding. If it’s automation, it’s still happening too slow in contrast to what shoppers want. In the rare case where AI is actually being applied, yes there is cause for concern for both bias and data privacy. The trade-off between convenience and privacy is and will continue to be one of the big dualities when it comes to data and technology.

Gene Detroyer
Noble Member
3 years ago

The adoption of AI has nothing to do with the pandemic. As we have seen with all our technologies, the speed of adoption just gets faster and faster. AI is here and is improving at a fast rate. Those that do not adapt will be left behind.

AI is not perfect, but waiting for perfect will leave you waiting forever. AI does not have an end point any more than most any other data-based technology.

Lisa Goller
Trusted Member
3 years ago

While the rapid pace of AI adoption seems overwhelming, the pandemic revealed the urgent need for new efficiencies.

With so many market factors moving at once, companies need a new approach. AI is versatile and effective at saving companies time and effort, so they can reallocate resources from repetitive tasks to helping customers. Integrating AI into business processes also helps retailers and brands spot and act on evolving consumer and competitive trends to boost agility.

Having the bandwidth and budget to manage all those AI-related factors is arguably the biggest hurdle.

Dave Wendland
Active Member
Reply to  Lisa Goller
3 years ago

Your identification of the biggest hurdles seems to be spot on, Lisa. As other panelists commented, I believe AI is here to stay but the real question is are organizations ready to take action?

Matthew Pavich
3 years ago

The reality is that AI adoption isn’t moving fast enough. In today’s complex, highly evolving retail landscape, AI is absolutely required to drive better, data-informed decisions that can drive key business objectives. Some of the concerns cited can actually be ameliorated through AI (an example being that AI pricing can actually result in more ethical outcomes through more efficient enforcement of anti-discriminatory pricing rules). The biggest barrier to adoption will always be change management and the ability of an organization to embrace new technology while understanding the delicate balance between science and the “art of retail.” One thing is clear, the retailers who are adopting AI and using best practices today will be tomorrow’s winners.

Brandon Rael
Active Member
3 years ago

If anything, considering the rapid pace of change due to the great digital acceleration, AI adoption is happening too slowly. Customers enjoy the personalization that comes with a good digital commerce experience. However, to achieve personalization at scale, we as a nation and our culture must understand and overcome any privacy issues that go along with sharing our personal data.

Our society is ready for this level of data sharing, as long as there are trust and transparency between the brands and the consumers. There are clearly valuable insights and experiences that could be driven through data and analytics. The modern-day consumer is more than happy to share these in exchange for an outstanding cross-channel experience.

We should expect that the bond and relationship between the retailers and brands and the customers will help move AI adoption to the next level.

Dave Wendland
Active Member
3 years ago

Interesting discussion and a real dilemma for some. The pace of AI and machine learning has accelerated and shows no sign of slowing. Is this a good thing? Yes, if a retailer or supplier can put the data to use and drive results. And, no. It’s not ideal if reliance on AI overlooks the requirement that systems, operations, and human resources must be aligned.

AI is definitely a frontier that needs to be explored and deployed. However expecting AI to solve underlying organizational issues or to shore up an unsteady foundation is a road to potential disaster. As my late father would say, “If the house is not in order, a fresh coat of paint and new furniture are merely temporary fixes.”

tpmichaud
3 years ago

I think AI adoption is retail is actually moving too slow. Tearing down data silos and understanding where companies can truly move the needle with AI will be central to its ultimate success. What is AI doing to generate profitable revenue growth? Part of the challenge for retailers is to cut through all of the noise associated with AI to understand what’s real AI and what’s not. There is no shortage of software vendors claiming to be AI who are actually not. AI Pinocchios if you will. At the same time, we see too many underfunded companies trying to enter the space. I have no doubt that there will be a vendor shakeout here like in other innovation areas. The good news is that AI does promise to transform many retail processes especially in marketing, merchandising, digital and supply chain.

Andrew Blatherwick
Member
3 years ago

The results of this research are conflicting and confusing, which is not surprising when the term AI is used in such an all-encompassing manner. As with all technology, there are good safe and efficient uses of AI and there are also uses that are more questionable. The concern about data security is true of any retail IT and companies need robust security management to protect themselves. If their data is vulnerable, AI will not change that. To say it is moving too fast begs the question “what parts of AI technology are moving too fast?” If they are delivering true value to the retailer and not using or abusing personal data, then the answer is no. If they are abusing personal data the answer is probably yes.

Yogesh Kulkarni
3 years ago

For many retailers, AI fundamentally represents an opportunity for “decision automation.” If you have a fast changing marketplace, supply constraints and decreasing staff to deal with it, you have to adopt AI.

We are indeed in an exciting time where where cognitive AI (machine learning that can read, respond, write, drive) is now getting adopted for business problems that relate to inventory management, pricing, personalization for retailers. But AI adoption is easier said than done. Most of the solutions in the marketplace in retail are black-box, don’t give users the confidence or guide them to key decisions and tell them the risks and reward with the decisions driven by AI.

There is a definite need to drive more innovation in the “adoption frontier” to make AI mainstream in the retail business. AI today is more like — you have it, I have it, all of of us have it, but most don’t know how to use it. But given the pace of innovation, this is about to change rapidly.

Ananda Chakravarty
Active Member
3 years ago

The pandemic, while moving more ecommerce, has touched AI, but only peripherally. AI was moving ahead at a rapid clip prior to the pandemic, but retailers haven’t found many use cases or applications for AI. Demand forecasting, price forecasting, in-store assistants, chatbots, etc. have been the going AI concerns. With less traffic to stores and more conscious customers, AI applications have slowed down, not quickened.

Retailers have been cautious about implementing new options and many have spent extra cash on barriers, masks and curbside pickup. The innovations implemented are mostly tied to digital and this is where the 47% of execs in the survey that suggest AI hasn’t quickened due to Covid-19 have figured most AI adoption has been seen.

The retail market remains overwhelmingly physical, even with the continued adoption of digital, hence the speed of change for AI is not rapid nor profound.

As for hurdles, data understanding is the key to AI, and all the others from ethics and bias to security have only scratched the surface of disruption. The AI space is not yet big enough within retail. When it comes to being retail specific, capabilities such as search engines or data access influence areas broader than retail issues. These are tied more closely to digital process, and aren’t explicitly retail hurdles. The bias aspect is something to watch however, especially as AI continues to grow.

Liza Amlani
Active Member
3 years ago

Working in retail merchandising for 20+ years, I can say that the adoption of ANY technology is way too slow. The year of digital transformation has been great in a sense of putting the need for retail technology and digitization at the forefront of a retailer’s survival.

The biggest hurdles that retail is dealing with now are that data is being stored in too many places and insights cannot be extracted, the use of legacy systems are literal blocks to digital transformation, and PEOPLE.

Leaders need to push new ways of thinking and working — change management must include humanizing the adoption of technology, such as AI, and shifting mindsets to look at AI as a necessary tool to keep retailers open.

AI needs to be implemented much quickly across demand planning and data analytics, to help merchants make data-driven decisions — we are currently moving at a snail’s pace. Let’s get this show on the road already!

Kenneth Leung
Active Member
3 years ago

The pandemic is driving increasing use of data because transactions and interactions are pushed virtual and therefore generating large amounts of data. There are several hurdles as identified such as inherited bias, data quality and security, but the technology is moving fast because of returns on investments potential can be so high.

Mohamed Amer
Mohamed Amer
Active Member
3 years ago

Technology adoption is a funny thing. Early adopters are excited to try it; laggards will fight it all the way to a bankruptcy filing. AI is unique because it’s similar to a new computing model, rather than introducing a new device or a release of a software application. AI challenges the way we think about business models, systems, and processes, even how we engage our customers and our supplier relationships.

Is AI adoption moving too fast? No! If anything, we are moving at a snail’s pace.