Artificial Intelligence In Retail:
What Now?
Benchmark Report
Brian Kilcourse and Steve Rowen, Managing Partners
October 2022
Sponsored by:
Executive Summary
Key Findings
Today, despite years of hype, confusion, and downright misinformation, practical concerns about
AI/ML are very much on the minds of retailers. New sources of data are available to businesses
from an increasingly digitized global marketplace, and the hope is that AI/ML technology can be
used to turn those data into insights, and help retailers better understand the environments they
operate in. As a result, we set out to study retailers’ attitudes about these very real-world
opportunities.
The following are some of the highlights of what they told us:
• Nearly half of our retail respondents chose the ability to identify trends that impact future
objectives as their #1 use for business analytics, despite 12 different choices on offer. In
an age of increased uncertainty, retailers can’t rely on gut feel to know what to do next,
and they want the smartest math possible (AI) to be able to help them react. Quite simply,
they want (and need) better tools to navigate uncertainly than what they currently have.
• The best performers, (those whose sales are already outperforming the norm) are FAR
more bullish on AI and ML tools. 71% of the best performers say AI-enabled analytics
will fundamentally change how they forecast their merchandise in the next three years
(compared to only 44% of average and underperformers). This pattern extends to all kinds
of decision-making processes, including supply chain planning and management (62% vs
48%) and store performance evaluation methods (56% vs. 40%).
• While retailers’ number one external challenge continues to be rapid shifts in consumer
demand (46% choose it as their top priority from a list of 9 options), the biggest opportunity
they see for greater use of AI-enabled analytics turns inward: 44% say they can use these
advanced algorithms to help them react better to supply chain interruptions. Quite simply,
retailers fear consumers – and rightfully so. But they think their best chance of upping their
game (again, correctly) is to improve their own ability to react to the unthinkable. At all
costs, they know they at least need to be able to provide the products they’ve promised
that they can. No doubt, the COVID-19 pandemic has shaken our industry up, and AI-
enabled analytics may well be one of the biggest benefactors of that dramatic change.
• As it relates to specific technologies, our response pool tells us that investment is coming
slowly: but it is, indeed, coming – particularly for the best performers. Retailers are not
moving as quickly to leverage these new tools as consumers might wish, but slow-and-
steady increases in both the value and use of such AI-enabled interfaces as executive
dashboards, exception reports and visual data models has steadily improved since we last
conducted this research during the height of the pandemic in 2020. It is the ability to layer
intelligence onto the structured data they are collecting from their legacy operational
systems that holds their interest most: both right now, and in the coming future.
Based on our data, we also offer several in-depth and pragmatic suggestions on how retailers
should proceed. These recommendations can be found in the Bootstrap Recommendations
portion of the report.
We certainly hope you enjoy it,
Brian Kilcourse and Steve Rowen
ii
Table of Contents
Executive Summary ........................................................................................................................... i
Research Overview ......................................................................................................................... 1
Fast-Paced Adoption Of A Revolutionary Technology ................................................................ 1
Flexibility As Strategy ................................................................................................................... 1
How AI Fits In ............................................................................................................................... 3
Retail Winners And Why They Win .............................................................................................. 4
What New Data? .......................................................................................................................... 5
Methodology................................................................................................................................. 7
Survey Respondent Characteristics ............................................................................................ 8
Challenges On Both Sides Of The Model .................................................................................... 9
The Kitchen Sink ........................................................................................................................ 10
The State Of Current Merchandising And Marketing Analytics ................................................. 12
Opportunities ................................................................................................................................. 14
Top Of Mind ............................................................................................................................... 14
How To Improve Supply Chain Management? .......................................................................... 15
What About The Consumer Side Of The Business? ................................................................. 17
Who Benefits? Everybody! ......................................................................................................... 18
Organizational Inhibitors ................................................................................................................ 20
The Plot Thickens ...................................................................................................................... 20
Pictures Bring Stories To Life .................................................................................................... 21
Most Retailers Still In ‘Education’ Mode .................................................................................... 22
Technology Enablers ..................................................................................................................... 24
Time In A Bottle ......................................................................................................................... 24
Making Legacy Tech Smarter .................................................................................................... 26
Change Is Gonna Come ............................................................................................................ 27
BOOTstrap Recommendations ..................................................................................................... 30
Don’t Fear The Reaper .............................................................................................................. 30
Use These Tools To Help Do What You Said You Could ......................................................... 30
Think About The Merchandising Opportunities .......................................................................... 30
Think Outside The Box .............................................................................................................. 31
Think Inside The Box, Too ......................................................................................................... 31
Take A Page From Winners’ Book ............................................................................................ 31
Appendix A: The BOOT Methodology
©
........................................................................................... a
Appendix B: About Our Sponsor ..................................................................................................... b
Appendix C: About RSR Research ................................................................................................... c
iii
Figures
Figure 1: Greater Flexibility Is The Strategy .................................................................................... 2
Figure 2: Still Early Days ................................................................................................................. 4
Figure 3: Winners Are Clear-Eyed ................................................................................................... 5
Figure 4: New Data To Feed New Analytics ................................................................................... 6
Figure 5: Sensing Demand / Managing Supply ............................................................................... 9
Figure 6: More Differences Than Agreement ................................................................................ 10
Figure 7: The Need To Improve Store-Level Execution With Insights .......................................... 11
Figure 8: A Lot Of Upside For Current Analytical Capabilities (Part 1) ......................................... 12
Figure 9: A Lot Of Upside For Current Analytical Capabilities (Part 2) ......................................... 13
Figure 10: A Reflection Of The Real World ................................................................................... 14
Figure 11: It’s All About Visibility ................................................................................................... 16
Figure 12: Fashion Lags ................................................................................................................ 16
Figure 13: Going Real-time ........................................................................................................... 17
Figure 14: The Whole Company Could Benefit ............................................................................. 19
Figure 15: Help Wanted ................................................................................................................. 20
Figure 16: A Picture’s Only Worth A Thousand Words When It’s Mobile ..................................... 22
Figure 17: Still ‘Early Days’: But For How Long? .......................................................................... 23
Figure 18: As Time Passes, Need Only Escalates ........................................................................ 24
Figure 19: Getting It Done ............................................................................................................. 25
Figure 20: Grabbing The Low Fruit First ....................................................................................... 25
Figure 21: The Aperture Widens ................................................................................................... 26
Figure 22: Now We’re Talking! ...................................................................................................... 27
Figure 23: What Will Be Important… ............................................................................................. 28
Figure 24: …What’s Actually Getting Done ................................................................................... 29
1
Research Overview
Fast-Paced Adoption Of A Revolutionary Technology
Artificial Intelligence (AI) and Machine Learning (ML) technologies have been the subject of a lot of
hype in the past several years. For some, AI/ML has the potential to trigger a dystopian future
where automation does everything from completely replacing the human workforce, to chatbots
instead of humans to answer our questions, all the way to Terminator-like "killer robots". It doesn't
help that some of the most respected thinkers in the world such as theoretical physicist Stephen
Hawking (“artificial intelligence could spell the end of the human race”) and Elon Musk (who once
referred to artificial intelligence as "summoning the demon"), have expressed grave concerns.
Meanwhile in the retail industry, much more practical concerns about AI/ML are on the minds of
retailers. There is a veritable ocean of “external” data available to businesses from an increasingly
“digitized” global marketplace, and the hope is that AI/ML technology can be used to turn those
data into insights, and help retailers better understand the environments they operate in.
RSR noted in its 2020 benchmark on retailer attitudes about AL/ML that retailers are trying to
change their operational models to be able to respond very quickly to sudden changes in either
supply or demand, and they are looking to AI/ML to help them to achieve that. What was new to
this quest is the application of AI technology. But retailers also recognized that moving from a
hyper-efficient (albeit not very flexible model) to a hyper-agile one is a huge change that affects
employees and business processes, as well as the technologies that support them.
The purpose of this new benchmark study is to update our understanding of if and how retailers
are focusing AI/ML to address challenges in the marketplace and seize on the opportunities that
those challenges create.
Flexibility As Strategy
Retail is a famously reactive business; when either supply or demand changes, retailers
change in response. RSR’s own BOOT research methodology reflects a 12-24 month
horizon in recognition that retailers rarely look too far beyond the next year’s plan. Retailers’
planning cycles assume a predictable environment, and the strategic imperative is operational
efficiency. But in recent years (and certainly since the Great Recession of 2008-12), the retail
marketplace has been buffeted by a seemingly endless procession of mega-challenges:
economic and political uncertainty, new and powerful competition, the rise of digitally
empowered consumers, pandemics, and global climate change.
In short, the world in which retailers operate isn’t predictable, it’s increasingly dynamic. That
in turn is causing businesses across industries to adopt agility as a strategic imperative, i.e.,
developing an intrinsic ability to respond very quickly to changes in the marketplace.
Retail is certainly not immune to this business trend. We can see this in the priority that retailers
assign to data analytics (Figure1). Just as it was in 2020, retailers want to be able to identify
trends that can affect their ability to meet future objectives.
2
Figure 1: Greater Flexibility Is The Strategy
Source: RSR Research, October 2022
As we’ll see later in this report, retailers are anxious to use new data gathered from outside the
theoretical four walls of their businesses to develop scenario models – with the expectation that
they will be able to better position the business’s processes and assets to respond much more
quickly to changes in the marketplace that may occur. Greater flexibility is the strategic objective.
This is very different from the tactical objectives that retailers were focused on just ten years ago.
RSR’s 2012 study on the state of analytics in retail
1
revealed that the top opportunity for retailers
at the time was to improve their ability to match demand with assortments, prices, and promotions.
Retailers sought to use new customer sentiment data from the then-emerging digital selling channel
as well as from the market-basket to optimize their merchandise plans.
But retailers haven’t forgotten about operational efficiency as they seek greater agility– narrow
profit margins never let retailers ignore the need to optimize internal processes. That is why the
top-three uses for data include the ability to optimize operations and to put actionable (i.e., real
time) information into the hands of operators.
1
Retail Business Intelligence: A Work in Progress, RSR Benchmark Report, October 2012
!"##
##$%
&'##
(""
)
)
&'
)* )
+"
,-" "
'
." "-/
What Are The TOP THREE (3) Most Important Uses For
Data Analytics Within Your Company?
3
Agility and optimization aren’t mutually exclusive. In this study, we wanted to see how retailers view
AI’s ability to help them achieve both objectives.
How AI Fits In
In RSR’s 2020 study, we noted that over-performing retailers were far more bullish on AI-analytics
to better manage products, the supply chain, and even to evaluate store performance. Based on
our observations, we identified several to-dos for retailers:
• Accept the “meta challenge” and use AI-infused analytics to help rapid responses and
decision making;
• Use AI infused merchandise forecasts and plans to put the right product in the right place
at the right time;
• Expand decision automation to be able to support greater localization of merchandise
planning and execution;
• Use AI to glean insights from non-transactional data;
• Support Store Operations with AI;
• Don’t abandon traditional analytics; AI is not a “silver bullet”.
All of these recommendations are predicated on retailers having the ability to develop AI models
(or “algorithms”) to make sense of all the new data available. We wanted to test that assumption,
and so we asked retailers to self-assess their AI “readiness (Figure 2). While it’s very clear that
retailers don’t believe that AI is “a solution looking for a problem”, there is no question that AI
adoption is still in its early stages, at least as evidenced by how far companies are in developing
the internal capabilities needed.
4
Figure 2: Still Early Days
Source: RSR Research, October 2022
Within these findings, there are some interesting insights. For example, over-performing “Retail
Winners” (described below) are more likely to have data scientists on board (40% compared to
25% of all others). And curiously, far fewer Fashion & Specialty retailers than others have data
scientists in-hand (only 6%), although far more of those retailers have committed budget to
acquiring that talent (59%, compared to 33% overall). Clearly, they are late to the AI game.
It's also interesting to note that the self-assessment provided by our survey respondents is almost
identical between line-of-business managers and IT’ers. The message is that business decision
makers and technologists are in agreement about where their companies stand.
Retail Winners And Why They Win
Over the years, RSR has found significant differences between retailers who over-perform in year-
over-year comparable sales and their competitors. These differences are highlighted in our
benchmark reports; consistent sales performance turns out to be an outcome of a differentiating
set of thought processes, strategies and tactics. While some might argue that comparable sales
are a dated metric, it remains the best measure of retailer success.
RSR’s definition of “Retail Winners” is straightforward. Assuming industry average comparable
store/channel sales growth of 7 percent in 2021, we define those with sales above this hurdle as
“Winners,” those at this sales growth rate as “average,” and those below this sales growth rate as
“laggards” or “also-rans.”
0
1
23)"
)4.5--"
6 7 *"-8#
5
2*-
9 *"+#- 3
2 - ")*"
4.:;-
2 !%-
9
2 -
)5 -
Thinking About Your Company’s AI Readiness, Choose The
Statement That Most Closely Describes Your Position Now
(Choose ONE)
5
Figure 3, below, gives us our first insight into the differences in thought processes and execution
capabilities between Retail Winners and their peers.
Figure 3: Winners Are Clear-Eyed
Source: RSR Research, October 2022
While average and under-performers are hedging their bets about the potential impact of AI,
Winners are both more bullish about the new technology’s potential for impact on both demand
forecasting and supply chain planning, and more transparent in their assessment that they don’t
really know the true extent of AI’s impact to internal operations in the future.
More Winners than non-winners are also willing to challenge some of the hype surrounding AI.
That’s encouraging; As Apple’s Steve Jobs once said, “It's not the tools that you have faith in - tools
are just tools. They work, or they don't work.” After it is all said and done, AI alone won’t make a
business better – retailers can’t wave it like a magic wand. People make businesses better. But
Winners also know that AI is an important new tool that will help their businesses turn data into
insights – and those insights will help to make the business better.
What New Data?
Figure 3 shows a majority of Winners see value in the power of AI-enabled analytics to help them
improve demand forecasts and merchandise plans, and to improve supply chain planning and
management. Those are clearly the top focuses even for non-winners. But most Winners also think
that AI-enabled analytics will help them to get the best performance out of their stores and improve
their interactions with customers.
0
1
0
0
1
6 "4.5--)
)4.5-)
) +
4.5-) )
) +
4.5-)" )
" +
4.5-) """
+
4.5-)" )
)""
+
Please Rate Your Reaction To The Following Statements
('Strongly Agree')
2 &
6
That’s all good news. The question is, what data will feed those AI-enabled analytics? Here’s what
retailers told us (Figure 4):
Figure 4: New Data To Feed New Analytics
Source: RSR Research, October 2022
Interestingly, retailers are more united in their assessment of the value of new data than they are
in the role of AI to help them make sense of it all. As a reflection of the findings in Figure 3, supply
chain alerts and sudden changes in key consumer data are ranked as “very important” by a large
majority of retailers (and particularly by Retail Winners).
Some high priority data are what we would expect: competitor information and location
demographics/psychographics, in conjunction with new data like commute patterns and business
schedules, could help retailers plan store locations, localized fulfillment centers, and merchandise
assortments and presentations.
Other data, such environmental data, real time social incidents, and community health, all point to
retailers’ desire to be more proactive in how they respond to conditions that could impact their
ability to execute day-to-day operations effectively.
Looking at these results by retail vertical, there are a few stand-out learnings. For example, FMCG
(“fast moving consumer goods”) retailers see more value in environmental data than all other
retailers (78% compared to 73% overall), while GM (“general merchandise”) retailers are far more
concerned about real time social incidents than all other retailers (90% compared to 69% overall).
These are reflective of the peculiarities of those particular verticals (for example, FMCG’ers’ ability
to bring fresh food to the sales floor is impacted by the weather).
0
0
1
1
1
1
10
10
11
<
=5!#
#%
< ! ""#$%
>
;
<"
? - ##
(!) #$%
*
=5 !+ #
-*##$%
Rate The Importance Of The Following New Sources Of
Data In Your Decision-Making Processes ('Very Important')
2 &
7
But taken as a whole, retailers of all stripes want to use new data to gain new insights – they only
seem to disagree on whether AI-enabled analytics is a prerequisite to being able to accomplish
that.
In the following sections of this report, we will identify how retailers are addressing the business
challenges and opportunities associated with AI-enablement. To a great extent, this is a Winners’
story. But as we’ve already seen, it’s still early days for AI-enablement, and non-winners have a
chance to catch up. Throughout the report, we’ll point out if and how they are trying to leapfrog their
way into a better position.
Methodology
RSR uses its own model, called The BOOT Methodology
©
to analyze Retail Industry issues. We
build this model with our survey instruments. See Appendix A for a full explanation.
In our surveys, we continue to find the kinds of differences in thought processes, actions, and
decisions cited above. The BOOT helps us better understand the behavioral and technological
differences that drive sustainable sales improvements and successful execution of brand vision.
8
Survey Respondent Characteristics
RSR conducted an online survey in August 2022 and received answers from 100 qualified retail
respondents. Respondent demographics are as follows:
• Products Driving The Majority Of Revenue:
Fast Moving Consumer Goods (C-store, Food &
Drug, Health Care Products)
45%
Apparel, Footwear, & Specialty 14%
Hard Goods (CE, Hard Goods, Home Décor,
Improvement, Automotive)
18%
General Merchandise (Discount, Mass Merchant,
Department Stores)
20%
Vertically Integrated Brand 3%
• Retail Presence:
USA 100%
Canada 43%
Latin America 22%
UK 21%
Europe 20%
Middle East & Africa 6%
Asia/Pacific 8%
• Year-Over-Year Sales Growth Rates (assume average growth of 7%):
Worse than average 11%
Average 37%
Better than average (“Retail Winners”) 52%
• Respondents Position Within The Organization
C-level (e.g., CEO, CFO, COO, CIO) 46%
Vice President 17%
Director/Manager 35%
Staff and other 2%
• Functional Area of Responsibility
Executive Management 36%
Customer Experience/eCommerce Operations 5%
Finance, Legal & HR 8%
Information Technology 37%
Merchandising 3%
Marketing 4%
Store Operations 1%
Procurement, Supply Chain and Other 6%
9
Business Challenges
Challenges On Both Sides Of The Model
The retail operational model has three basic components: the consumer side, the supply side, and
all the operational processes that bring those two sides together. We have already noted that
Winners are bullish about the new technology’s potential for impact on both demand forecasting
and supply chain planning. In the aggregate response, we see these concerns clearly reflected
(Figure 5).
Figure 5: Sensing Demand / Managing Supply
Source: RSR Research, October 2022
But other concerns are also highlighted; competition, consumers, and (as is always the case)
operational costs. These concerns come into sharper focus when we look at differences by
performance.
It’s hardly surprising the supply chain disruptions are top of mind for Retail Winners and non-
winners alike (Figure 6). 2021-22 saw shocking interruptions in the flow of goods to retailers. As
RSR noted in its 2021 benchmark study on the state of the retail supply chain:
“When it comes to greater visibility into the supply chain, especially “alerts for unexpected
supply shortages” and “critical inventory situations anywhere in the supply chain”, over
60% of retailers believe those to be a problem. Finally, only 52% of our survey respondents
are confident that they can “respond to disruptions… Taken as a whole, the overall picture
<+'
+
<3@ @A
) -
<-7) 8
<- )
*
<-78#
"""
2
55)*$
<""
"
2--?
-9*
< #
---)
TOP THREE (3) Business Challenges Your Company Faces
That Create Interest In Expanding The Use Of AI-Enabled
Analytics
10
is that of an industry that has difficulty anticipating, seeing, or responding to disruptions in
the supply chain whenever they occur.”
2
Figure 6 shows that Winners and non-winners agree on one key point, that they need to be able to
detect supply chain disruptions quickly enough to take corrective actions.
Figure 6: More Differences Than Agreement
Source: RSR Research, October 2022
From there, however, concerns quickly diverge. A majority of non-winners believe that sudden
shifts in consumer demand are undercutting their ability to lower cost of goods through efficient
buying practices. On the other hand, Winners worry that somehow “the competition” is nimbler than
they are.
We’ve seen this before now. In RSR’s 2022 benchmark on the state of KPIs in retail
3
, we noted
that 56% of Winners (compared to 33% of other retailers) identified that “the competition is much
more agile in responding to changes in supply and demand than we are”. In that report, we noted
that “over-performers stay sharp by acting as if the competition is catching up.”
The Kitchen Sink
To better understand if and how next-generation analytics can help retailers optimize the internal
processes that bring supply and demand together, we asked retailers to prioritize operational
challenges that potentially could benefit from improved insights.
When it comes to store operations, what we got was a veritable “kitchen sink” full of operational
issues, but three stand out: the need for more consistent store-level execution, the growing problem
of returns handling, and customer demands for more services that add new costs (Figure 7).
2
Retail Supply Chain: Navigating Through Rough Waters With Improved Agility, RSR Benchmark, December 2021
3
How Retailers Are Operationalizing Analytics With New KPIs, RSR Benchmark, June 2022
0
0
<-78#
"""
2
55)*$
<""
"
< #
---)
<- )
*
2--?
-9*
Top Business Challenges That Create Interest In AI-Enabled
Analytics
(Selected Differences)
2 &
11
Figure 7: The Need To Improve Store-Level Execution With Insights
Source: RSR Research, October 2022
Let’s address the potential role of insights derived from analytics for each of these challenges:
First, retailers are looking to improve store operations by informing operational processes with
insights derived from real time analysis of business metrics. RSR’s 2022 benchmark study on the
state of KPIs in retail
4
identified many of the insights derived from improved analytics that could
help. They include labor retention rates and utilization rates, sales per employee, inventory turns
and out-of-stock measures, shrink and waste measurements, and customer order fulfillment
measures.
Secondly, as relates to in-store returns of online orders, RSR recently revealed in our annual report
on the state of the stores
5
that retailers see the ability to accept online order returns in the stores
as an opportunity as much as a challenge, since it creates another chance to interact with
4
How Retailers Are Operationalizing Analytics With New KPIs, ibid.
5
What Can Retailers Do In Stores That Amazon Still Can’t?, RSR Benchmark, August 2022
0
0
1
0
0
&)"3
"!*#
#%
<-)5-
-")
&+
&""""
.)
-+
2B --
"
2*) ")
& ""
"9)-
<+ "
)-
. -""
)
C"+:
TOP THREE (3) Operational Challenges Associated With
Your Company’s Current Business Analytics Capabilities
12
consumers in the store. Retailers understand the role of analytics in helping them to control this
increasingly important activity.
Finally, retailers recognize that in order to service customers in new ways, they need to “find the
money” by optimizing and even automating some in-store decisions and processes. That requires
operational analytics.
The State Of Current Merchandising And Marketing Analytics
Analytics has had an important role to play in bringing “science” to the art of retailing. Along with
financial analysis, merchandising and marketing functions have been using data warehouses and
related tools for over 30 years to understand product movement by location, time period, and by
assortment categories. Vendor performance is score carded, and supply chain performance is
measured. Promotional effectiveness is monitored, and customer interactions are tracked to
improve loyalty. These capabilities are accomplished largely with traditional BI (“business
intelligence”) tools that examine internal data from operational systems, and don’t require the new
data identified earlier in Figure 4.
Figures 8 and 9 highlight retailers’ self-assessment of their current analytical capabilities as they
relate the supply side and the customer side of the business. Two realizations jump out: first,
Winners make better use of analytics, and secondly, even Winners are challenged to get more from
those capabilities.
Figure 8: A Lot Of Upside For Current Analytical Capabilities (Part 1)
Source: RSR Research, October 2022
0
0
0
11
4
4-
,5 "
&' "- :*
." -
."5+
D)" #
Rate Your Company’s Current Merchandising And Supply
Chain Business Analytics Capabilities
('Full Capability' - Selected Differences)
2 &
13
Figure 9: A Lot Of Upside For Current Analytical Capabilities (Part 2)
Source: RSR Research, October 2022
Under-utilization of current tools amounts to a true business challenge for retailers – even for Retail
Winners. Adoption of AI to gain new insights from new data has great promise, but many companies
aren’t getting the full value of their current capabilities. RSR’s already mentioned 2022 benchmark
study on the state of KPIs in retail
6
gave us some clues about why this is the situation. They include
siloed data, “dirty” or incomplete data, and “too many versions of the truth”.
While we’ll find out more about what stands in the way of AI adoption in the Organizational
Inhibitors section of this report, there’s a more practical to-do facing retailers right now, and that
is to use the tools and data that are already available.
6
How Retailers Are Operationalizing Analytics With New KPIs, ibid.
0
0
0
0
0
0
0
1
&'
? """#"-
&'
"*
&'
,' +
."7-8:"-
Rate Your Company’s Current Marketing And Other
Customer Facing Business Analytics Capabilities
('Full Capability' - Selected Differences)
2 &
14
Opportunities
Top Of Mind
Retailers can be forgiven if their primary focus is the supply chain. Many are still hurting from the
massive disruptions experienced in 2021. For example, retailers like Minneapolis-based Target
attributed its weaker-than-expected first quarter results to “inventory impairments” and costs
relating to supply chain disruptions. Very recently, S&P Global Chief U.S. Economist Beth Ann
Bovino commented that “It's still a major problem… It's one of the biggest factors that are causing
this where we are today. We have seen some signs of softness, some signs of moderation — but
nowhere near what we need to get to."
Earlier in this report, retailers told us that one of their top business challenges is the need to detect
supply chain disruptions and react as quickly as possible (Figure 6). It follows then that the top
opportunity they see is the other side of that challenge: to improve reaction to supply chain shocks
(Figure 10).
Figure 10: A Reflection Of The Real World
Source: RSR Research, October 2022
Looking at the details of that finding, something interesting is revealed: while improved reaction to
supply chain shocks is the top opportunity for both Retail Winners and all others, it is non-winners
that put the most weight on it (50% compared to 38% of Winners). But that doesn’t imply that
0
0
0
0
> )*
E-") #
- - "
""
F#*
) B
.-"5
'"
-
>-"'
>@) "@-"
##
4- "
-
. "
. *
TOP THREE (3) Opportunities From Greater Use Of AI-
Enabled Analytics Within Your Business?
15
Winners don’t see that as an opportunity. Rather, it’s likely that most Winners are already pretty
good at reacting to disruptions in the supply chain. Fully one-half of average and under-performers
rate that to be a top opportunity- an indication that those retailers are not satisfied with their current
capabilities.
There’s a similar dynamic in the second top opportunity, to improve reaction to sudden shifts in
consumer demand. In Figure 4, average and under-performers identified rapid changes in
consumer demand as their top business challenge. In the responses to the question in Figure 10,
44% of non-winners view improved reaction to sudden shifts in consumer demand as a top
opportunity, compared to only 35% of Winners. Although that’s a narrower gap than the opportunity
for improved reaction to supply chain shocks, it still points out that more Winners are confident that
they can respond appropriately.
How To Improve Supply Chain Management?
In RSR’s 2022 benchmark on the state of IoT (“Internet-of-Things”) in retail
7
, respondents identified
the Supply Chain as the top opportunity for impact from deploying IoT. IoT and AI are closely related
in that they work together to deliver real time process insights to decision makers. In the IoT report,
we noted that:
IoT creates a lot of non-transactional data … new data tools, particularly artificial
intelligence (AI) and machine learning (ML), make it possible for retailers to observe and
measure the effect of their efforts in real time and to alert operators when exceptions occur.
In this study, retailers identify several capabilities that are aided by real time insights (Figure 11).
7
A Deep Dive Into Retailers’ Views About RFID And The Internet Of Things. RSR Benchmark, March 2022
16
Figure 11: It’s All About Visibility
Source: RSR Research, October 2022
A majority of respondents rated each of the opportunities we presented as “high value” (with
Winners leading the way). But when we looked at responses by retail vertical, far fewer Fashion &
Specialty retailers assign a high value to some of these supply chain-related insights (Figure 12).
Figure 12: Fashion Lags
Source: RSR Research, October 2022
Earlier in this report, we noted that Fashion & Specialty retailers lag behind the other retailers in
bring data scientists in-house to develop AI capabilities. At least as it relates to supply chain
management, fewer of these retailers see a lot of value to real-time insights. These findings are
0
0
1
0
1
1
10
10
11
1
.")*-*'
59
=
."""
D--55) )
."-
F"
D- ""
Rate How Real-Time Insights Can Help The Following
Supply Chain-Related Tasks
('A Lot Of Value')
2 &
1
1
0
1
1
.")*-*'
59
F"
."""
D--55) )
Rate How Real-Time Insights Can Help The Following
Supply Chain-Related Tasks:
Selected Differences ('A Lot Of Value')
4;; #?#>
17
related, and probably are indicative of the long supply chains associated with niche and design-to-
sell private label product assortments.
What About The Consumer Side Of The Business?
This report makes the connections between new data, AI-enabled analytics, and real time insights.
These technologies are all supportive of a digital transformation strategy. Realtime visibility is a key
objective of any digital transformation strategy; any physical “thing” that has a digital equivalent can
be observed in real time to determine its status and analyzed.
While the supply chain is a primary focus for many retailers’ digital transformation agendas, there
are many opportunities on the consumer (selling) side of the business as well. We asked retailers
to rate the value of real time insights for operational processes on the consumer side of the
business, and just as on the supply chain side, we found that Winners are leading the way (Figure
13).
Figure 13: Going Real-time
Source: RSR Research, October 2022
There are some interesting differences by retail vertical. While retailers generally agree that real
time insights will be an aid to merchandise planning & forecasting, and cross-channel order
management. In the case of merchandise planning & forecasting, only FMCG retailers (grocery,
drug, convenience) assign a lower value than the overall group, and more General Merchants
assign ‘high value’:
0
0
0
0
0
0
1
0
1
1
1
1
10
10
1
<5
?
,5'
?
(-
?""
.*
2--
"
,5")
Rate The Value That Insights From New Sources Of Data
Could Bring To The Following Operational Processes At Your
Company ('A Lot Of Value')
2 &
18
74;&"D8 <E E
#
?#
>
G
"
In the case of customer order management, General Merchants and Hardgoods retailers express
the greatest interest while fewer FMCG and Fashion & Specialty retailers assign ‘a lot of value’:
74;&"D8 <E E
#
?#
>
G
<5 0 0 1 1
But beyond these exceptions, more Retail Winners assign a ‘high value’ to real time insights for all
the customer facing capabilities we asked about. Clearly, real time insights constitute a winning
behavior – something that all retailers should aspire to.
Who Benefits? Everybody!
Finally, we asked retailers to tell us which corporate functions would benefit from real time insights.
The answer? All of them (Figure 14).
19
Figure 14: The Whole Company Could Benefit
Source: RSR Research, October 2022
Surprisingly, there was general agreement both by performance groups and by retail vertical.
Beyond any of the detailed responses, this speaks to retailers’ perception that they need to
constantly monitor their operations in the face of a fast paced and ever-changing environment.
In RSR’s 2022 benchmark on KPIs
8
, we noted that “There is strong agreement that management
is “constantly looking for new ways to measure performance” and that “the executive team needs
performance reports that are ‘short and sweet’”. AI-empowered analytics are seen as a way to get
those new measures.
In the next section of this report, we’ll identify what stands in the way of retailers’ ability to address
challenges and seize opportunities.
8
How Retailers Are Operationalizing Analytics With New KPIs, ibid.
57%
60%
65%
66%
71%
71%
73%
74%
75%
76%
82%
1
;,
#;#G=
,F
*
?&
<:F&
."6 !.6%
?<
<(+
(+
How Much The Following Departments In Your Organization
Can Benefit From Getting Insights In Near Real Time From New
Sources Of Data?
4;&"D ?D ;6CD
20
Organizational Inhibitors
The Plot Thickens
When we compare what stands in the way of addressing the opportunities that come from greater
use of AI-enabled analytics in comparison to the last time we conducted this research, the answer
is simple: as retailers have learned more about AI, they have developed a more sober assessment
of what stands in the way of adoption (Figure 15).
Figure 15: Help Wanted
Source: RSR Research, October 2022
1
1
1
0
&@@:
&+
@"-@*
6 -@@"
6 ") -"
>
4-"
2B- )9*
"
> - "
4.5
& 9 )
9
.B 9=&.")4.5-
-
&.69)*) )
+
2 " )
5"
###
TOP THREE (3) Organizational Inhibitors Standing In The
Way Of Taking Advantage Of Opportunities
21
Retailers don’t have a whole new host of internal roadblocks to overcome - quite the contrary. Their
problems have simply heightened. Their number one challenge is still that they need help
understanding how they can use all of the new data they are collecting, specifically the non-
transactional data that is “new” to them.
For example, what are retailers to do with a sentiment expressed on Facebook? Determining
context, alone, is still remarkably challenging – even Facebook’s own hyper-advanced AI tools still
can’t even get that part accurate. And the idea of making intelligent decisions based on the troves
of data streaming in from search engines can be overwhelming. Retailers know they need help,
more than they even did a few years ago.
While it’s encouraging to see fewer retailers report budgetary constraints, that can only be
expected; the last time we conducted this research in 2020, AI enablement was still a new topic for
many retailers. As a result, the fact that so few retailers had earmarked budget for AI technologies
in the summer of 2020 was hardly surprising.
Pictures Bring Stories To Life
As mentioned in the Business Challenges section of this report, RSR recently concluded its first
ever study on next- generation key performance indicators, How Retailers Are Operationalizing
Analytics With New KPIs
9
. One of the main takeaways from that report was retailers’ opinion that if
new analytics are to mean anything, they need to be visually interesting: particularly to executives.
From that report:
Our retail respondents are clear: in order for the impact new KPIs enable to reach maximum
effectiveness, they must be presented in a manner that works best for key decision makers.
Clean and simple layouts are paramount. Line of business executives are often-times quick
to dismiss overly technical data presentations (regardless of how impressive they are or
how much work went into them), and therefore dashboards that make huge swaths of
complicated data inherently actionable hold the most favor.
Today, in Figure 16, below, retailers echo this sentiment. When it comes to AI, highly visual
presentations for decision makers are the best way to get past the vast array of hurdles they have
defined.
9
How Retailers Are Operationalizing Analytics With New KPIs, ibid.
22
Figure 16: A Picture’s Only Worth A Thousand Words When It’s Mobile
Source: RSR Research, October 2022
It is surprising to find was that providing decision makers with mobile access to those compelling
visuals was equally important as the graphics, themselves. But our respondents are clear: big
decisions are being made by executives at any time of day. If the benefits of AI are to be leveraged,
they not only need to make the results of complex analytics easy to consume, but they also need
to be available via a mobile device anytime and anywhere there’s Wi-Fi or cellular connection.
This is a big deal.
Most Retailers Still In ‘Education’ Mode
When asked to choose the most pertinent statement to describe their efforts to date, retailers
indicate that it is still early days for AI adoption: more reporting being in the “educational” phase
than any of the other options we offered (Figure 17).
1
1
4"4.
:5
<"
&"4.5-
(+
6 )
)'
?
G)
G!5#@@
-#$%
, ")
4.5-4 9
-4.5 "+
G "
*! -#$%
TOP THREE (3) Ways To Overcome Organizational
Inhibitors
23
Figure 17: Still ‘Early Days’: But For How Long?
Source: RSR Research, October 2022
Once more, good news comes in small doses. While most retailers have at least one AI project in
pilot, we had expected to see more further along the maturity curve. In all likelihood, retailers have
far better ideas of ways AI could help than their investments to date would suggest. In fact, they’ve
been telling us that very thing since our earliest forays into AI inquiry back in 2017. When it comes
to advanced use cases? Of course, they could use more education from the industry about the
possibilities. But as we are about to see in the Technology Enablers section of this report, the jury
has long been out on some very real, very right-now functions that benefit from the analysis and
intelligence that AI solutions bring to the table.
Perhaps the brightest spot in Figure 17, however, it is that very few retailers (only 10%) are currently
doing “nothing” with AI-enabled analytics. That means it is time to find out exactly they are doing
with these next gen technologies.
0
2B 4.5-
2 "
/ 4.5-#
*)/
F""'#
#4.5/#-
" -
2 #4.5-
/ -
"""$
2 "
4.5-#-
2
!/%
What Is Your Company's Current Level Of Experience
When It Comes To AI-enabled Analytics?
24
Technology Enablers
Time In A Bottle
As we’ve just seen in the previous section, retailers say the way to get past corporate aversion to
new, AI-enabled analytics and tools is to provide executives with visually compelling dashboards.
It only makes sense, then, that their importance is growing with the passing of time (Figure 18).
Figure 18: As Time Passes, Need Only Escalates
Source: RSR Research, October 2022
Far and away, dashboards top the list of what’s not only important right now, but what will be
important in the future. Nearly 8 out of every 10 retailers define these tools as important to their
strategy going forward.
But the critical role new tools will play is not limited to dashboards that make complex analytics less
daunting. Exception alerts, newly redesigned KPIs, the ability to search as simply as one can in a
web browser: all of these interfaces have grown in importance in the past 2 years. What’s just
important as their perceived value, however, is their growing use over that same amount of time
(Figure 19).
0
0
01
0
01
0
0
01
0
1
C;."
&978!7-*8#$%
=5--
7E5*8
(+
D
FH,.
F -
'Very Important' Interfaces
25
Figure 19: Getting It Done
Source: RSR Research, October 2022
In just the past 24 months, the entire industry has elevated its use of these tools. Better still, they
are happy with their investments, to date. Who, then is driving such positive trends?
As we can see in Figure 20, the answer is Retail Winners.
Figure 20: Grabbing The Low Fruit First
Source: RSR Research, October 2022
0
1
0
0
(+
D
=5--
C;."
7E5*8
F -
'Implemented & Satisfied'
0
0
0
0
0
0
(+
FH,.
7E5*8
D
&978!7-*8#$%
=5--
C;."
F -
What Is The Status Of The Following Interfaces In Your
Company?
('Implemented & Satisified')
2 &
26
Retail Winners didn’t become winners by accident. While much of the industry talks about the future
value that artificial intelligence will bring to commerce, they have been harder at work incorporating
the tools they know will make the most sense first. To wit: more than half of Retail Winners have
not only brought on every interface that we put forth as an option, but they are already happy with
each of those solutions. That’s a big deal.
From the ability to engage via spoken word (natural language interfaces) to knowing when things
aren’t going quite to plan (exception alerts) - and everything in between - Winners are seeing
benefits from the investments they’ve made so far. While some of these certainly classify as “low
hanging fruit”, they are wins, nonetheless.
Making Legacy Tech Smarter
When our line of inquiry changes to the legacy capabilities that retailers have in house, we find a
significant opportunity for retailers to up their games (Figure 21).
Figure 21: The Aperture Widens
Source: RSR Research, October 2022
Nearly 3 out of 4 retail respondents cite structured data extracted from their operational systems
as the capability they still rely on most. On its own, the ability to make this data more actionable is
a clear opportunity for AI. Delivering new insights from structured data is exactly what will enable
retailers to stay competitive in a world increasingly dominated by Amazon.com, by enabling smarter
order orchestration, enhanced understanding of customer patterns and behaviors, even the ability
to understand and manage product returns.
But as it stands, Figure 21 reminds us that much of this analysis happens using legacy capabilities.
1
00
00
0
0
1
1
8F)8-
="7"8-
=5-
4--!#)*# #
9%
F'-,#;#6
?
F')
F+:":"""
?+"
How Important Are The Following Legacy Capabilities To
Your Company?
(- . (- :;.
C(- :I. C(- :IC.
27
Sixty-three percent of our respondents say that that data is still analyzed with spreadsheets. This
is hardly what comes to mind when one thinks of the future of retail – but as is so commonly said
these days: it is what it is. The good news is that nearly 30% of retailers say this process is losing
importance. In a similar vein, nearly half of our respondents identify drill-down capabilities as in use
and valuable, but nearly 40% say that capability is falling out of vogue.
In short, retailers have increasingly mixed feelings on which of the things they’ve been doing for
years are now worth doing the same way. And that signifies something worth calling out: even if
retailers are slowly warming to the concept of AI as a real-world asset, they are – quite clearly –
beginning to question their old processes and tools. Change cannot come without this critical step,
and this data clearly shows that for the majority of retailers, this is the stage they are in.
Change Is Gonna Come
As is so often the case, however, the best performers show us why they are Winners. In Figure 22,
Winners show us what matters most to them, and as we’ve learned to expect, they are much further
down the road than their average and underperforming peers in several key areas.
Figure 22: Now We’re Talking!
Source: RSR Research, October 2022
Winners, just like everyone else, still do a lot of analysis with spreadsheets. But they place a much
higher priority on structured data (nearly 20%), and the same can be said for their implementation
of real-time and store data aggregations.
What’s more, Winners have different ideas about which data analytics capabilities stand to benefit
them most in the foreseeable future (Figure 23).
0
0
0
0
0
1
1
1
8F)8-
=5-
F'-,#;#6
F')
4--!#)*# #
9%
="7"8-
F+:":"""
?
?+"
'Established & Important'
2 &
28
Figure 23: What Will Be Important…
Source: RSR Research, October 2022
As Figure 23 shows, Winners’ response to which analytics hold most value is “yes”. They place
more value on math-based analysis (21% more than average and underperformers), what-if-
analysis (by a measure of 15% more), and when it comes to video analytics, their appetite is nearly
30% greater than that of their peers. They just want more.
It should come as little surprise, then, that when we look at the status of these same capabilities,
Winners are vastly more engaged in the types of analytics that they say hold the key to success in
the future (Figure 24).
0
0
0
0
0
0
0
0
0
1
1
1
1
10
10
1
?-) 4.
,
D4
?.6) 4.
,:
C;9
?!2 5."%
7E5*8
7 5-8! %-
= 9
How Important Are The Following New Data Analytics
Capabilities To Your Company?
('Very Important')
2 &
29
Figure 24: …What’s Actually Getting Done
Source: RSR Research, October 2022
Quite simply, the future is brighter for those who are already doing best. Retail Winners have gotten
to where they are by looking at challenges and opportunities differently, and AI adoption is no
exception. Are they moving as quickly as they could? Perhaps not. But they are moving towards a
future where AI-infused analytics help assist in all kinds of decisions at a rate faster than average
and lagging retailers are.
And as it turns out, that’s fast enough.
Now let’s make some recommendations based on all we’ve learned from these findings.
0
1
0
0
0
0
0
0
0
0
C;9
?-) 4.
?.6) 4.
= 9
,
D4
?!2 5."%
7 5-8! %-
,:
7E5*8
What Is The Status Of Each Of The Following Data
Analytics Capabilities In Your Company?
!B.:?"B%
2 &
30
BOOTstrap Recommendations
As with every report we conduct, RSR concludes by offering some baseline recommendations that
all retailers can keep in mind as they continue on their paths – in this case, as it relates to their AI-
enabled journey. This is a unique field of inquiry for us, and due to its growing importance, one we
may choose to fold into every topic within retail we study going forward (Merchandising, Supply
Chain, The Store, etc.).
With that in mind, here our recommendations for how retailers should be thinking about AI in the
meantime:
Don’t Fear The Reaper
While no one can 100% guarantee that mankind’s ultimate demise won’t come at the hands of
robot overlords operating via AI protocols that have begun thinking for themselves and are now too
powerful to stop, one thing is for certain: those quasi-sentient beings won’t be trying to determine
the best price set for a pallet of soy milk rapidly approaching its sell date.
AI – for the purposes of the retail industry – is just another tool in the toolbox. One that is really
powerful, and can help retailers make even smarter decisions – about a whole raft of things they
have to decide every day – than they currently do. Staffing decisions, merchandising decisions,
supply chain decisions: all stand to benefit. The time for AI in our industry is well at hand, and
retailers need to start acting faster to leverage its benefits - before the real nemesis (their
competitors) do.
Use These Tools To Help Do What You Said You Could
COVID-19 shook the retail industry to its core. The wildly fragile nature of the just-in-time supply
chain was revealed to anyone watching (including shoppers), and – seemingly overnight – the
“supply chain” became conversation among people who had likely never heard the term before
2020. As a result, one of the things retailers continually tell us in all of our recent research (this
project, included) is that they need to improve their reaction to supply chain shocks. Customers
are, in essence, demanding it. They have quickly grown intolerant for out-of-stocks (either online
or in-store) and are always one Amazonian click away from leaving a brand who can’t deliver what
it said it could. Don’t be that brand. Retailers need to utilize every tool at their disposal to keep
Amazon.com from becoming the shopper’s “default mode.” AI holds tremendous opportunity to
combat that very thing by helping retailers to precisely deliver the right products to the right places
at the right time and in the right quantities to meet local demand – without over-inventorying.
Think About The Merchandising Opportunities
For many years now, Over ten years ago, RSR analyst Paula Rosenblum sounded the alarm about
a “sea of sameness”: the phenomenon brought about by many (if not most) retailers selling
remarkably similar products. We all understand how we got here: the goal is to provide enough
ubiquitous products so as to eliminate shoppers’ need to visit multiple stores.
The problem is that when too many retailers are selling the same “stuff,” what makes one store
more attractive than another has relatively little to do with the product mix. And when nearly
everything a retailer sells can also be bought at a big box store, the chances to compete become
far slimmer. The need for differentiated products has never been greater. Can AI really help
31
determine what some of those products might be? Can it help with creating a differentiated mix?
Winners think the answer to these questions is “yes”. We do too.
Think Outside The Box
There’s no shortage of ways to leverage the teachings that can come from math-based analytics.
While many retailers may not have been early to the AI party, decision makers should take some
time outside of the normal routine to think about what processes could be improved by smarter,
math-based analyses. Each brand is different, so while it is difficult to recommend that people carve
out the time to “think differently”, that is precisely what we are doing. Retail decision makers need
to think about their brand, what makes it different from the competition, and if there’s anything that
could benefit from a “smartening up” of that process. The more unique that “it” is, the better.
Think Inside The Box, Too
Retailers that operate stores should consider all the ways math-based analyses could help allocate
any (or all) of the things that happen within the four walls. Staffing requirements by store, season,
and time of day, employee training and scheduling, product pricing – all could benefit enormously
by algorithmic analysis of not just historical trends, but also “what if” scenarios. Winners are
disproportionately interested in scenario-based testing and are well underway in pilot programs
with many such store-focused solutions. This is not the time to let them pull away beyond the point
of being able to catch up.
Take A Page From Winners’ Book
Lastly, those who engage with an RSR analyst for any length of time will invariably hear the phrase
“Winning is not an accident.” It’s not. It’s an outcome that results from looking at the world in a more
curious and less staid manner. And Winners are no different when it comes to AI.
It’s easy – even tempting – to dismiss artificial intelligence as the stuff of science fiction, the stuff
of “one day.” It’s not. AI is already helping the best retailers across the world rationalize what they
buy and in what quantity, how to get those products where they need to be, where they will sell
best and at what price – the list goes on and on. Winners are embracing the potential. Everyone
should.
a
Appendix A: The BOOT Methodology
©
The BOOT Methodology
©
is designed to reveal and prioritize the following:
• Business Challenges – Retailers of all shapes and sizes face significant external
challenges. These issues provide a business context for the subject being discussed
and drive decision-making across the enterprise.
• Opportunities – Every challenge brings with it a set of opportunities, or ways to
change and overcome that challenge. The ways retailers turn business
challenges into opportunities often define the difference between Winners and
“also-rans.” Within the BOOT, we can also identify opportunities missed – and
describe leading edge models we believe drive success.
• Organizational Inhibitors – Even as enterprises find opportunities to overcome their
external challenges, they may find internal organizational inhibitors that keep them
from executing on their vision. Opportunities can be found to overcome these
inhibitors as well. Winning Retailers understand their organizational inhibitors and
find creative, effective ways to overcome them.
• Technology Enablers – If a company can overcome its organizational inhibitors it can
use technology as an enabler to take advantage of the opportunities it identifies. Retail
Winners are most adept at judiciously and effectively using these enablers, often far
earlier than their peers.
A graphical depiction of the BOOT Methodology
©
follows:
b
Appendix B: About Our Sponsor
SAS is the leader in advanced analytics. Our solutions help enhance merchandise & assortment
plans, drive deeper customer insight, better plan for demand, and optimize inventory & pricing.
That’s why over 1,400 retailer and consumer goods companies worldwide – including the top 7
global Consumer Goods - companies rely on SAS.
Visit www.sas.com/retail for more information.
c
Appendix C: About RSR Research
Retail Systems Research (“RSR”) is the only research company run by retailers for the retail
industry. RSR provides insight into business and technology challenges facing the extended retail
industry, providing thought leadership and advice on navigating these challenges for specific
companies and the industry at large. We do this by:
• Identifying information that helps retailers and their trading partners to build more
efficient and profitable businesses;
• Identifying industry issues that solutions providers must address to be relevant in the
extended retail industry;
• Providing insight and analysis about a broad spectrum of issues and trends in the
Extended Retail Industry.
Copyright© by Retail Systems Research LLC • All rights reserved.
No part of the contents of this document may be reproduced or transmitted in any form or by any means without the
permission of the publisher. Contact research@rsrresearch.com for more information.
