THE FUTURE OF RETAIL 2018: ARTIFICIAL INTELLIGENCE
P R O V I D I N G I N - D E P T H I N S I G H T , D A T A , A N D A N A L Y S I S O F E V E R Y T H I N G D I G I T A L
THE FUTURE OF RETAIL 2018:
ARTIFICIAL INTELLIGENCE
DANIEL KEYES, RESEARCH ANALYST
$994
$1,374
$1,900
$2,628
$3,635
$5,034
2017 2018 2019 2020 2021 2022
Estimated revenue for companies providing AI to retailers, millions ($)
Source: Research and Markets, 2017
Spending by retailers on AI is expected to reach
$5 billion by 2022
Source: Accenture
AI is projected to boost
profitability nearly 60%
for retail and wholesale
by 2035 +59%
profitability rates
by 2035
AI’s future in retail is the brightest in three areas:
1. Personalization
2. Search
3. Chatbots
Brands that use
personalization see sales
rise by up to 10%
Source: Boston Consulting Group, 2017
+6%-10% revenue increase
2-3 times faster than those that don’t use
personalization
Note: Seven percent of all e-commerce visits included a customer clicking on a recommendation.
Source: Demandware, 2017
Consumers who click on a recommendation are more likely to
convert and spend more
Share of orders and revenue from visits where consumers clicked on a recommendation
Orders Revenue
24% 26%
Percent of shoppers who made a return visit to a site
Source: Demandware, 2017
And they are almost twice as likely to make
a return visit
37%
19%
Clicked a recommendation during first visit Did not click a recommendation during first visit
MACHINE LEARNING
A branch of artificial intelligence that
automates the building of analytical
models, allowing a machine to act
without being programmed
Enables large-scale data
analysis that can be
used in predicting
consumer trends and
behaviors
Machine learning crunches large data sets to
identify trends
And uses those trends to
maximize positive outcomes
from consumer interactions
Machine learning can help create personalized online
interfaces that make the most of consumers’ time
This can help retailers
reach impatient mobile users
quickly and efficiently
Source: ContentSquare, n=300 million sessions, 2017
Which is key, as most
mobile sessions with retailers
are less than a minute long
And because mobile is driving
e-commerce growth
US m-commerce sales volume
Source: Business Insider Intelligence estimates, US Census Bureau, comScore
$35 $55 $76 $107 $137 $177 $226 $286 $359 $447
12%
16%
20%
23%
25%
28%
31%
33%
36%
39%
2014 2015 2016 2017 2018E 2019E 2020E 2021E 2022E 2023E
Billions ($) Percentage of e-commerce sales
The AI identified regional trends where consumers hesitated or
repeatedly clicked on the page
Source: ContentSquare, 2017
Russian users hesitated
13 seconds longer on the shipping page
than the world average to modify their billing address
Allowing L’Occitane to modify regional web pages,
boosting sales on mobile
Source: ContentSquare, 2017
+15% m-commerce sales
Results like these are
why 72% of retailers
plan to invest in
machine learning
Retailers planning to invest in machine learning/cognitive
computing by 2021
Source: Zebra Technologies, n=1,700 retailers, 2017
72%
Search is crucial to the shopping experience because 80% of
shoppers regularly use site search
Source: RichRelevance, n=1,033 US consumers, 2018
80%
always or often use site search while
shopping
And 72% will abandon a retail site that produces poor search
results
Source: RichRelevance, n=1,033 US consumers, 2018
72%
are likely to leave a retail site that
doesn’t provide good search results
NATURAL LANGUAGE
PROCESSING
The process that enables computers to
understand human language
•Improves search results
•Chatbots
•Tracks and tags customer
sentiment via social media
Natural language processing (NLP) is a type of AI that can be used
to improve search results
NLP can map items to conversational
words and phrases
COMPUTER VISION
Technology that allows
computers to obtain
information from images
• Visual search
• Gathers customer behavioral
data in stores and tracks wait
times in lines when combined
with cameras and sensors
• Enables self-checkout by
sensing when items are
picked up off shelves
• Augmented reality try-on
features
Computer vision powers visual search, letting users search based on
pictures they take or find
Global number of digital photos taken on mobile devices, in billions
Source: Mylio
Visual search taps into consumers’ growing preference for
taking photos
764
894
1020
2015 2016 2017
Source: Pinterest
Which has helped Pinterest Lens, a visual search tool,
rack up 600 million searches a month
250
600
Feb-17 Feb-18
Monthly Pinterest Lens visual searches, millions
Source: Pinterest, February 2018
Connecting shoppers to products from
all types of retailers
Pinterest Lens’ top categories for visual search:
• Fashion
• Home decor
• Art
• Food
• Products
• Animals
• Outfits
• Beauty
• Vehicles
• Travel
80% of consumers say most
of the time they try to resolve
an issue online before
contacting customer service
Source: Radial & CFI Group, n=500, 2018
Percentage of US consumers who stopped doing business with a company because of bad
customer service
Source: Aspect, n=1,000, 2018
And 54% of consumers stopped doing business with a company
due to bad customer service
49%
54%
All shoppers
2016 2017
Percentage of US consumers who stopped doing business with a company because of bad
customer service
Source: Aspect, n=1,000, 2018
And that number climbs to 61%
among millennials
49%
53%54%
61%
All shoppers Millennials
2016 2017
Top reasons consumers would stop using a product, service, or brand
Source: Forrester & American Express, n=1,027 North American consumers, 2017
Two increasingly important issues
Gen Z cares about
16% 17%
9% 10%
23%
20% 21%
12%
Poorly designed mobile features Slow responding online chat for
sales or customer service
issues
Poor features/responsiveness
on social media
Customer service only available
via phone or in person
Gen Y (23-37) Gen Z (16-22)
NATURAL LANGUAGE
PROCESSING
Enables computers to understand
human language
•Chatbots
•Improves search results
•Tracks and tags customer
sentiment via social media
NLP and machine learning are the types of
AI behind chatbots
MACHINE LEARNING
Automates the building of analytical
models, allowing a machine to act
without being programmed
•Enables large-scale
data analysis that can
be used in predicting
consumer trends and
behaviors
AI can be difficult to implement
due to a lack of usable data,
difficulties with storage systems,
and expenses
57%
57%
61%
63%
64%
Belief that human
judgment is superior
to AI
Resistant to change
Employees
concerned about job
safety
Security and data
privacy
Lack of skills and
talent
B i g g e s t c h a l l e n g e s t o i m p l e m e n t i n g A I a m o n g c o m p a n i e s t h a t h a v e
d o n e s o
Source: Capgemini, n=939 companies, 2017
And because AI talent is
extremely scarce