In this talk, I explore the question of whether true autonomic marketing possible. To answer this question, I provide an overview of AI applications in advertising and marketing with an emphasis on digital marketing. Specifically, I survey the applications of AI in promotions, advertisements, search, pricing and programmatic marketing. Given the current state, I then explore what might be possible in the future. Towards this, I present some of the trends in AI uncovered by Omnicom Media Group and our point of view on AI applications in advertising and retail. Finally, I conclude with some case studies of applying AI at OMG and lessons learned.
2. Overview
• What is autonomic marketing?
• Retail evolution
• Applications of AI in marketing and advertising
• What is possible in the future?
• Omnicom Media Group POV
• Where are we in the journey
• Case Studies
• Chatbots
• Visual Listening
• Lessons Learned
Overview
3. What is autonomic (or
algorithmic) marketing?
“a marketing process that is automated
to such a degree that it can be steered
by setting a business objective in a
marketing software system”
- Introduction to Algorithmic Marketing
“That’s where marketing is headed. We
are on the verge of the need for
autonomic responses just to stay afloat.
Personalization, recommendations,
dynamic content selection, and dynamic
display styles are all going to be table
stakes.”
- AI for Marketing
Source: Wikipedia
4. Technology has evolved retail, from cash registers to robot
concierge
RetailEvolution
1960
First electronic cash
registers are invented
2016
Facebook messenger and Skype
launch bot platforms, Amazon Go
opens, Hilton pilots robot
concierge, Google Home launches
2015
Amazon Echo launches in USA
2014
Apple Pay is launched,
Android and Samsung
Pay follow a year later
2010
Introduction of Pinterest
& Instagram
1967
Automated teller machines
are invented in London
1969
Chemical Bank installs the first
ATM in the US, Interactive voice
response (IVR) technology
develops – but the first systems
were complicated and
expensive
1974
Barcodes begin the
data revolution
1981
In the 1980s IVR technology
becomes affordable for more
companies, Thomson Holidays UK
is the first business-to-business
online shopping system
1984
The self service till is
invented by David R
Humble – inspired by
standing in a long
grocery checkout line
in South Florida
1991
Internet becomes
publicly available
2009
Bitcoin is released
and WhatsApp is
founded
2008
Androids and un-
preinstalled apps
began appearing
2007
iPhone
2006
Twitter
2005
YouTube
2004
Facebook
2001
iTunes allows customers to
purchase individual songs online
1998
Google and PayPal
are founded
1997
Jeeves the butler
uses Natural
Language on
Ask.com
1996
Hotmail launches
1995
QueueBuster,
the first
intelligent call
back solution,
provides
customers with
an alternative to
spending hours
on hold
1995
eBay launches
1994
The birth of ecommerce
with Amazon launching
Source: AI@OMD
5. AI in promotions
• Response, LTV modeling
• Targeting
• Promotion optimization
• Design of promotional creative
ApplicationsofAIinmarketingandadvertising
Source: ILYA KATSOV (2018), Introduction to Algorithmic Marketing:
Artificial Intelligence for Marketing Operations
Targeting
Server
Data
Management
Platform
Analytics Platform
Request
Response
Objective
Use data and models
Load models
6. AI in search
• Optimization of sponsored search
• Visual search
• Voice search
• Local search
• Amazon and Apple search
ApplicationsofAIinmarketingandadvertising
7. AI in pricing
ApplicationsofAIinmarketingandadvertising
• Dynamic pricing and pricing optimization
• Competitor pricing
• Introductory price optimization
• Aggregate pricing optimization
Source: ILYA KATSOV (2018), Introduction to Algorithmic Marketing:
Artificial Intelligence for Marketing Operations
8. AI in programmatic marketing
• Programmatic advertising
• In-depth behavioral customer segmentation
• Highly targeted automated media buying
• Example: Albert AI and Harley Davidson NY
• Programmatic CRM
• High level customer personalization
• Dynamic audiences
• Real-time triggers
• Cross-channel reach
• Example: Blueshift CRM and OYO India
• Retargeting
ApplicationsofAIinmarketingandadvertising
Source: James Cannella (2018), Artificial Intelligence in Marketing
9. What is possible in the
future – Experts' Opinions
• Marketers are able to specify high-level campaign goal
metrics
• Marketers have a way to measure brand awareness
• Media optimization becomes machine-driven
• The all-in-one intelligent marketing hub emerges
• An open-ID alternative emerges
More intelligence in the marketing
systems reduces the need for workflow
and enables marketers to focus on more
strategic work.
Source: CHAVEZ, O’HARA, VAIDYA, Data Driven (2019): Harnessing Data and AI
To Reinvent Customer Engagement
10. Omnicom Media Group Point of View
Whatispossibleinthefuture
MARKETING
WITH
PROGRAMMATIC
CS
INFLUENCER
ENGAGEMENT
MARKETING
OF
SERVICE
EXPERIENCES
CAMPAIGNS
VOICE CAMPAIGNS
MARKETING
TO
SEO
VOICE SEARCH
VRM
Developing quickly
Driving performance
improvement
The long term vision
Outsourced consumer
decision making
Today’s emerging
challenge
Key to success in the
long term
Source: AI@OMD
11. Where are we in the journey?
Whatispossibleinthefuture
There are 5 stages of AI. Currently, we are progressing from interactivity with AI to co-activity
Source: AI@OMD
12. Chatbots
CaseStudies
EASY MOMENTS NEED-TO-KNOW MOMENTS IMMERSIVE
MOMENTS
RECEIVE REWARDS FROM
YOUR FAVOURITE BRAND:
GET UP TO SPEED WITH
GAME OF THORNES:
TRY ON DIFFERENT
MAKE-UP LOOKS:
STEP INTO A TV SERIES:FIND THE PERFECT
GIFT:
Source: AI@OMD
13. Visual Listening
CaseStudies
Data
#Love
Men & Women
Analysis
Other hashtags
image
insight!
SEARCH
MEN &
WOMEN
CONTEXT
Scrape photos with #love. Use social network analysis (SNA) to detect
contexts in which people talk about love:
marriage, having fun together, couple goals,
weddings, etc.
Use AI to detect only
photos of couples 18+
who share their feelings
online
In partnership with
wspólna zabawa
ślub
małżeństwo
Source: AI@OMD, Annalect Poland
14. Lessons Learned
CaseStudies
AWARENESS CONSIDERATION PURCHASE EXPERIENCE RETENTION
AI suggests nearby points of
interest based on my preferences.
AI understands what I want and
makes personalised suggestions
AI shows me options based on
my social accounts & ‘likes’.
AI asks me for parameters.
AI takes care of scheduling
appointments/delivery.
AI connects me to
expert opinions.
AI provides me with
quicker customer service.
AI offers seamless
payment options. AI opens a direct channel
with brand after purchase.
AI connects me to an online
community for feedback.
Digital Touchpoints
Physical Touchpoints
AI provides me with relevant advice
based on my habits and characteristics.
AI makes it easier to find
opening times, stocks, etc.
AI provides in-store assistants with my
past activity to offer the best service.
AI assists on making more informed
decisions about my purchases.
AI surprises me with new
opportunities and discoveries.
There will be people who will default to a particular channel (phone, website, etc.) and others who will use
multiple touchpoints, ensure you curate a consistent experience
TAKE A H UMAN FIRST APPROACH
Think about your existing customer journey and how
AI can enhance different touchpoints
Source: AI@OMD
15. Is true autonomic
marketing possible?
Programmatic advertising and CRM
Data driven creative development and optimization
Novel interactions and experiences
− Promotions, Pricing
Mitigating risks to the brand
Proprietary models and methods
Elimination of bias
Multiple optimization levers
Killer Creative
“The machine is not going to envision a
product that will be embraced by
millions for a good long time to come.
The machine is not going to invent a new
slogan that speaks directly to the hearts
and minds of the masses for a good long
time to come. The power to raise
awareness, instill trust, and create desire
is still in human hands for the time
being. Data is going to have a tough time
beating killer creative”.
- Jim Sterne, AI in marketing
Vijay Dakshinamoorthy
Vijay.Dakshin@omnicommediagroup.com