The Smarter Store
How AI is powering the future of retail
The technology market
intelligence platform.
CB Insights software lets you predict, discuss and
communicate emerging technology trends using data
in ways that are beyond human cognition.
See what other customers have to say at http://www.cbinsights.com/customer-love
A few of our happy customers
The most publicly reference-able customers of anyone in the industry
Beti Cung, Corporate Strategy, Microsoft
"We use CB Insights to find emerging
trends and interesting companies that
might signal a shift in technology or
require us to reallocate resources."
@cbinsights
#AIretail
Join the conversation on Twitter
December 12 13, 2017 | San Francisco, CA
A-ha! is a gathering of Fortune 1000 executives, investors,
academics, and startup founders discussing the next wave of
ideas and technologies that will shape our world.
SPEAKERS INCLUDE
Daymond John
The Shark Group
Star of ABC's Shark Tank
and CEO
FUBU
Founder/CEO
Liza Landsman
Jet.com
President
Sold to Walmart for $3.3B
Mudassir Sheikha
Careem
Co-founder & CEO
MENA's $1.2B Unicorn
Benedict Evans
Andreessen Horowitz
Partner
"Working Out What's
Happening Next" at a16z
Deepashri Varadharajan
@deepashriv
dvaradharajan@cbinsights.com
2
About the presenter
Senior Intelligence Analyst and Writer, Artificial Intelligence
Deepashri Varadharajan received her undergraduate degree in
engineering at Vellore Institute of Technology in India. She later
studied journalism, and received a degree from the Columbia
University Graduate School of Journalism.
Before CB Insights, Deepashri worked for organizations including Al
Jazeera America, Deccan Herald, and interned at Siemens India.
Zoe Leavitt
@zoe_leavitt
zleavitt@cbinsights.com
3
About the presenter
Senior Intelligence Analyst, Consumer
Prior to joining CB Insights, Zoe worked in equity research and drove
consumer analytics efforts at the consumer finance group of the
Bank of China in Shanghai.
Her research has been cited in the New York Times, the Wall Street
Journal, and more. Zoe is a graduate of Stanford University.
1
2
3
4
5
Macro trends in retail
AI in e-commerce
Startup ecosystem
AI in the physical store
Trends to watch
9
The Smarter Store:
How AI is Powering the Future of Retail
Macro trends
Automation, AI reshapes retail
10
Amazon dominates e-commerce, with AI core
11
The e-commerce giant has a strong AI focus. Its internal projects include Lex (NLP),
Rekognition (vision). Prime customers can now place orders through Alexa.
More than 3,200 stores are shutting
down here's the full list
April 17, 2017
A 'retail apocalypse': Stores close at record rates
12
Adapt or die.
Retailers are partnering with startups to survive in a highly competitive landscape,
understand their customers, and drive traffic both online and offline.
13
How automation is transforming retail
14
MANUFACTURING
ROBOTS
WAREHOUSE
AUTOMATION
ONLINE
IN-STORE
Logistics &
moving items
Inventory
management &
demand forecasting
Pick & place
robots
Packaging
robots
Computer
vision for
measurements
& styling
Chatbots for
CRM
Personalized
product
recommendations
Visual
search
Targeted
ads
On & offline
consumer
behavior
analytics
Computer
vision for
automated
checkouts
Robots for
inventory
management
Robotic sales
assistants
Delivery &
logistics
Drones &
delivery
robots
Demand
forecasting
Route
optimization
AI in e-commerce
Use cases of artificial intelligence
15
Customers want to easily find what they're
looking for.
16
Alibaba Group backed Israel-based e-
commerce search startup Twiggle in Q2'16.
Vision: Finding items without keywords
17
Automobiles
Wine
General retail
General retail
Retailers want customers to never leave.
18
Dynamic Yield focuses on personalized search, recommendations, and
behavioral messaging, among other things.
Select investors: Baidu Capital, Bessemer Venture Partners, ClalTech, Global
Founders Capital, Innovation Endeavors, Marker and Vertex Ventures Israel
1:1 targeting and merchandising
19
Select investors: Bain Capital Ventures, Battery Ventures, Lightspeed
Venture Partners, New Enterprise Associates, and Salesforce Ventures
Retailers want to understand consumer behavior
across devices.
20
Appier uses deep learning to collect and analyze audience behavioral and
browsing data.
Select investors: AMTD Group Company, EDBI, Line Corporation, NAVER
Corporation, Sequoia Capital India, and SoftBank Group
AI in commerce
Startup trends, deals, top investors
21
Commerce is a leading industry for AI deals
22
https://www.cbinsights.com/research-artificial-intelligence-startup-deals
CATEGORY
Q3'12 Q4'12 Q1'13 Q2'13 Q3'13 Q4'13 Q1'14 Q2'14 Q3'14 Q4'14 Q1'15 Q2'15 Q3'15 Q4'15 Q1'16 Q2'16 Q3'16 Q4'16 Q1'17 Q2'17 Q3'17
HEALTHCARE
HORIZONTAL APPLICATIONS
CYBERSECURITY
COMMERCE
AD & MARKETING
BUSINESS INTL. & ANALYTICS
FINTECH & INSURANCE
IOT/IIOT
SALES & CRM
AUTO TECH
PERSONAL ASSISTANTS
HR TECH
EDUCATION
NEWS, MEDIA & ENTERTAINMENT
AGRICULTURE
HARDWARE FOR AI
TRAVEL
LEGAL
REG TECH
SOCIAL
PHYSICAL SECURITY
REAL ESTATE
IT & CLOUD SERVICES
SPORTS
SOFTWARE DEV. & DEBUGGING
MISC
Low
deals
High
deals
Global AI commerce equity funding since '13:
$1.26 BILLION
across
256 DEALS
23
Top 2017 rounds go to
startups in Asia
Top rounds in 2017
Trax (Singapore)
64M/Private Equity
Computer vision for retail
Appier (Taiwan)
$33M/Series C
Cross screen consumer
behavior analysis
24
AI commerce: Annual global financing history
2013-2017 (as of 11/617)
$59
$254
$223
$389
$334
$394
(Full-Year Run Rate)
28
40
56
71
61
72
(Full-Year Run Rate)
2013
2014
2015
2016
2017 YTD
Disclosed Funding ($M)
Deals
Q1'17 busiest quarter
for AI commerce deals
2017 opened with the
busiest quarter for deals,
followed by a slowdown
in the following quarters.
Outside the US, Q3'17
saw 7 deals to countries
in Asia, including 4 in
China, as well as 3 deals
to startups in France and
United Kingdom.
25
AI commerce: Quarterly global financing history
Q1'13 Q3'17
VC interest in AI
commerce peaked in
2016
VC and corporate
interest in AI commerce
picked up in 2015, when
investors like Google and
Salesforce backed
startups including Modsy
and Msg.ai. The trend
continued in 2016, before
a slowdown this year.
26
AI Commerce: Unique investors
2013-2017 YTD (as of 11/6/2017)
44
42
61
75
52
3
12
15
21
17
35
41
62
58
46
2013
2014
2015
2016
2017 YTD
Venture Capital
Corporate/Corp VC
Other
2017
Recent VC-backed deals
27
AI transforms the in-store experience
AI helps retailers a) turn massive amounts of available data into insights, and b) automate
rote tasks, freeing up store employees for higher-impact work.
28
29
Behind the scenes
AI can help manage repetitive, time-intensive tasks behind the scenes in stores, freeing up
store associates for higher-impact interactions with shoppers. Stores can become more
experiential and differentiated from e-commerce.
AI can analyze disparate
data streams to
optimize retail planning
Celect, spun out of MIT,
uses machine learning to
model customer
behavior and optimize
retailers' merchandising,
assortment, and pricing
strategies.
Celect raised $15.1M
from In-Q-Tel, Activant
Capital, August Capital,
Fung Capital
Management, and
others.
30
AI supports inventory management
31
Nextail uses AI to help retailers improve inventory
allocation and manage daily store replenishments.
Disclosed funding: $1.6M
Select investors: Nauta Capital
Blue Yonder uses AI to help retailers improve inventory
planning and reduce out-of-stocks.
Disclosed funding: $75M
Select investors: Warburg Pincus
Everseen uses machine vision and video analytics to
monitor the point of sale and reduce losses.
Disclosed funding: N/A
Select investors: Marcol
Eversight uses machine learning to help retailers and
CPG brands test and deploy better promotions.
Disclosed funding: $25.2M
Select investors: Bow Capital
32
Trax Retail combines AI,
IoT, augmented reality
Trax uses machine vision
to help CPG brands
monitor their products on
store shelves and create
detailed augmented
reality recreations of
store environments for
audits. It also works with
retailers to observe
stores in real time.
Trax raised $147.6M
from Warburg Pincus
and others, and in July it
acquired Nielsen's store
observation unit.
Select Clients:
Sources: Fortune, St. Louis Post-Dispatch, Recode, Progressive Grocer, CNBC
Walmart is Putting Shelf-Scanning
Robots in Over 50 U.S. Stores
October 26, 2017
Stores roll out robot staffers
33
SoftBank's humanoid Pepper robot already greets shoppers at 10K+ venues around the world.
Pepper can answer questions, interact via touchscreen, and collect shopper data. Pilot programs
reportedly boosted shopper traffic by 20-70%. However, if robots do not add clear value for
shoppers, the novelty may wear off and Pepper's impact on foot traffic may fade.
Humanoid Robots Attract Shoppers
34
Simbe Robotics and Fellow Robots produce in-store robots that scan shelves and monitor
inventory to reduce out-of-stocks, correct misplaced items and labeling errors, and gather analytics
for store management. Fellow works with Lowe's, Yamada, and others, and Simbe works with
Schnuck's, Ahold, and others.
While B2B Robots Scan Shelves and Inventory
35
Walmart plans to expand its partnership with Bossa Nova's robots to 50+ stores. The fully
autonomous robots use machine vision to scan shelves and monitor inventory. Bossa Nova has
raised over $24M in funding, and Intel Capital led its Series C in October.
Walmart Taps Bossa Nova Robotics for
Inventory Management
36
Beyond robots: could
holograms be on the
rise?
NY-based Satisfi, a
platform for
conversational
commerce, recently
partnered with
augmented reality
company VNTANA to
launch an interactive AI
hologram.
AI-powered holograms
could answer basic
questions for shoppers,
while observing store
activity and visitor
demographics.
37
Select Clients:
38
In-store engagement
Facial recognition and AI-powered digital screens can make stores more engaging, while
collecting shopper interaction data.
Sephora's smart mirrors
drive in-store
engagement
While Sephora has long
driven engagement by
offering in-store product
try-on, today it uses
screens to offer virtual
try-ons and makeup
tutorials.
These screens offer
memorable in-store
experiences, reduce staff
time spent on managing
try-ons, and collect and
record data about which
products shoppers try.
39
ModiFace technology
powers Sephora's
screens and app
ModiFace uses AI and
augmented reality to
power virtual try-on
experiences for Sephora
and other beauty brands.
ModiFace's patented
technology can identify
facial features and skin
conditions, then make
personalized product
recommendations or
show shoppers how they
would look using
different products.
40
Patent: Method, system, and computer program
product for generating recommendations for
products and treatments
Patent: System, method and application for
skin health visualization and quantification
Patent: Method and system for simulated
product evaluation via personalizing
advertisements based on portrait images
+
AI Startups Battle for Beauty Brands
41
Select Reported Clients:
Select Reported Clients:
Adaptive screens help
engage and monitor
shoppers
Advertima offers digital
screens that can analyze
in-store shoppers and
adapt content in real time
to engage them. It aims
to detect gender, age,
and other data, and will
allow shopper interaction
via speech, touch,
gesture, or smartphone.
Swiss-based Advertima
raised $5M from Migros,
Switzerland's largest
retailer.
42
43
Could facial recognition
replace payments?
Alibaba partnered with
Chinese startup Face++
to unveil facial
recognition-based
payments.
Face++ has raised
$608M from Foxconn
and others, and earned a
$1B valuation in Oct'17.
Alibaba is currently
piloting the pay-by-face
technology in select KFC
locations, and could
expand it to retail.
44
Automating checkout
Startups aim to use AI to automate the checkout process and take cashiers, and even cash
registers, out of the store.
QueueHop smart tags
help secure self-
checkout
QueueHop provides
connected security tags,
which unlock
automatically once the
shopper has paid at a
self-checkout kiosk or via
mobile app.
It aims to help retailers
speed checkout and
reduce cart
abandonment, while
providing AI-powered
personalized
suggestions to shoppers.
45
Select Clients:
Machine vision makes
first steps toward
checkout automation
Pre-launch startup
Standard Cognition aims
to use machine vision to
enable automated
checkout. The system
would track shoppers'
carts throughout the
store, and automatically
display the charges at a
self-serve kiosk by the
door. In the future,
shoppers could pay
automatically via mobile
app.
46
Numerous startups already operate fully automated, human-free convenience stores in China.
These modular stores integrate with mobile payment apps such as Alipay and WePay. Shoppers
scan their phones to access the store, then scan again to pay and exit. The stores aim to use AI
and machine vision to track shoppers throughout the process.
Human-free storefronts roll out across China
47
Xiaomai
Disclosed Funding: $36.9M
Select Investors: Sequoia Capital China,
Junzi Capital, Aplus Capital
BingoBox
Disclosed Funding: $14M
Select Investors: GGV Capital, Ventech China
F5 Future Store
Disclosed Funding: $5.9M
Select Investors: Sinovation Ventures, TCL
Capital
48
Managing the customer journey
As retailers face new opportunities to gather direct data from consumers online and offline,
they can use AI tools to turn that data into action.
Hundreds of startups
help retailers gather
offline data
Startups are using AI,
IoT, Bluetooth, and other
new technologies to
crack open the "black
box" of brick-and-mortar
stores. These new
observation methods
help retailers build a
clearer picture of offline
shopper activity.
49
50
And AI can help merge
online and offline data
to create better
customer profiles
Paris- based Teemo (dba
Databerries) tracks
shopper activity to
provide targeted offline
mobile marketing.
Its personalized
messaging aims to drive
in-store sales.
Teemo raised $17.9M
from Index Ventures,
Mosaic Ventures, and
others.
Select Clients:
AI helps brands track
and parse online
feedback
Startups such as
Quantifind ($49M in
disclosed funding),
Remesh ($3M), Revuze
($4M), and Converseon
($7.5M) use AI to help
brands monitor
consumer sentiment.
Others, such as Yotpo
and Signpost.ai, use AI to
drive shoppers to
contribute reviews.
51
52
How does AI support Walmart's strategy?
Walmart's patents focus on guiding shoppers through stores, speeding up the checkout
process, and keeping the shelves stocked.
AI supports omnichannel growth
Laurent Desegur
VP Customer Experience, WalmartLabs
53
Source: VentureBeat
"We are enhancing the shopping
experience through machine learning
we want to make sure there is a
seamless experience between what
customers do online and what they
do in our stores."
Sources: VentureBeat, RetailDive, Fortune, Engadget, Fortune
Walmart Expands Use of In-Store
Inventory Robots
October 30, 2017
Walmart builds and partners for AI growth
54
Improving e-commerce through AI
55
Sources: Predictive shopping patent, Machine learning system, Search term patent
Patent: Predictive shopping
Application: March 2017
Patent: Machine learning system
Application: October 2016
Patent: System and method for
calculating search term probability
Application: July 2017
Dozens of Walmart patents focus on checkout
56
Sources: Smart cart patent, Preauthorized payment patent, Item scanning patent
Patent: Smart cart for self-checkout of
retail merchandise
Application: November 2016
Patent: Distributed processing for
preauthorized payment
Granted: July 2017
Patent: Simultaneous item scanning in a
POS system
Granted: February 2017
Walmart has filed 25+ patent applications related to drones since 2016.
Observing and serving shoppers by drone
57
Sources: Dispatching a drone patent, Retail facility patent, Monitoring premises patent
Patent: Dispatching an airborne drone to
carry an item to a customer
Application: November 2016
Patent: Method to carry an item within
a retail shopping facility
Application: September 2016
Patent: Apparatus and method for
monitoring premises
Application: December 2016
58
What's Amazon doing in AI?
AI infuses Amazon's B2C and B2B businesses, while helping Amazon internally.
E-Commerce
Amazon focuses AI efforts on:
59
Logistics
Business Services
Dozens of patents aim
to automate supply
chain and logistics
By searching on the CB
Insights platform for the
keyword "unmanned"
within Amazon's patents,
we can pull up dozens of
automation patents.
60
Amazon considers AI for manufacturing
61
Patent: On demand apparel panel cutting
Granted: October 2017
AI and 3D printing create personalized products
62
Patent:
Fulfillment of
orders for items
using 3D
manufacturing
on demand
Granted: June
2017
Patent: Item delivery using 3D
manufacturing on demand
Granted: June 2017
Amazon automates the warehouse
63
Patent: Inventory facility
Granted: January 2017
Patent: Modular, multi-function smart storage containers
Granted: May 2017
Patent: Systems
and methods to
facilitate
human/robot
interaction
Granted: May
2017
AI supports entirely new warehouse methods:
Under water
64
And Above the Clouds
65
AI supports e-commerce leadership
Jeff Bezos
CEO, Amazon
66
"Machine learning drives our algorithms
for demand forecasting, product search
ranking, product and deals
recommendations,
merchandising placements, fraud
detection, translations, and much more."
In e-commerce, patents
improve search
AI helps Amazon
automate product
identification and image
tagging. This supports
faster categorization of
merchandise, and helps
the site pull up more
relevant products in
response to search
terms.
67
Patent: Identification of item attributes using artificial intelligence
Granted: October 2017
Amazon considers
enhanced security in the
face of AI
In several patents,
Amazon notes that as AI
improves, e-commerce
sites may have more
difficulty separating bots
from human users.
In response, Amazon
proposes advanced
CAPTCHA tests using
optical illusions, facial
recognition, and more, to
verify human users.
68
Patent: Turing test via reaction to test modifications
Granted: August 2017
Amazon's consumer smart home products use natural language processing and machine vision
opening new platforms for ordering products, while gathering new data streams about consumers'
at-home activity.
AI creates new purchase platforms
69
In a recent patent, Amazon proposes using at-home video hubs (similar to the Amazon
Show) to observe people's activities, emotions, biometric data, and more.
Combining sales and data collection
70
Patent: Automatically staged video conversations
Granted: October 2017
Amazon Show image
AI could power drone delivery
71
However, Amazon
struggles with human-
free stores
Amazon delayed the
launch of its much-hyped
automated grocery store,
Amazon Go.
Amazon Go opened a
pilot location in Seattle in
late 2016 for Amazon
employees, with plans to
launch publicly in 2017.
However, it failed to open
as scheduled, facing
problems in handling
store traffic.
72
Amazon productizes its
AI, serves other retailers
Amazon offers APIs that
let clients integrate
chatbots, voice
functionality, and image
recognition into products
without training their own
AI models.
It serves retailers
including Nordstrom,
Nike, Stitch Fix, and
Brooks Brothers.
73
Sources: CNBC, WSJ, CNBC, Business Insider, RetailDive
Walmart to Vendors: Get Off Amazon's
Cloud
June 21, 2017
AI & AWS: Retail support or threat?
74
Trends to watch
M&A targets, big players, future jobs
75
Future of retail is a fusion of two worlds
76
PHYSICAL
DIGITAL
AI
Startups are acquisition targets for brands and
big tech companies
77
Alibaba is spending $15 billion on researching
quantum computing, AI, and more
10/11/2017
E-commerce giants will continue pushing the
boundaries of artificial intelligence
78
Measurable impact and barriers to entry
Bill Lewis
EVP of Consumer Products (Retail and Distribution)
Capgemini Consulting
79
source: www.retailtouchpoints.com/features/trend-watch/despite-hype-don-t-expect-quick-adoption-of-robotics-in-retail
"The costs out of the gate for these
robots are high, especially to run
many of the tests. The use cases are
still being understood and I think right
now most retailers are trying to focus
on their e-commerce integration with
stores."
Preparing for a reshaping of the workforce.
80

If retailers look to develop tech in house, AI talent will continue to be in high demand.

In-store automation would eventually lead to fewer hands on deck for repetitive tasks.
But it would put the focus back on customer-facing tasks, strategic planning, and
product design, among other things.
Retraining employees is a recurring theme. But with advances of AI in e-commerce, retail
jobs like sales may not translate into e-commerce sales positions.
Where is all this data from?
The CB Insights platform has the details on all the
deals included in this report
81
CLICK HERE TO SIGN UP FOR FREE
CBInsights.com
@cbinsights
+1(212)292-3148