3. “The easier part is convincing the board to
fund an AI project. It’s the next step that
worries me: will we actually be able to
deliver and will our customers welcome
it…?”
F T S E 1 0 0 , D i r e c t o r o f M a r k e t i n g
4. Agenda
How technology is rapidly changing the
competitive landscape and making the need
to understand and meet customers’ needs
key to success
The catalyst
Some mechanisms businesses are using to
respond to the shifting bases of competition
and meet customer needs
How businesses can respond
AI as an example of technology influencing
customer experience
7 Guiding Ideas
Real world learnings on how to think about
implementing a customer-first strategy in an
AI- context, covering board-alignment,
execution, technology, people and
organisation
5. ‘We both step and do not step in
the same rivers’
H e r a c l i t u s , 5 3 0 - 4 7 0 B C
6. Today’s science is tomorrow’s next big technology
T h e c a t a l y s t
Source: Gartner
7. On whether Netflix was a threat
“is the Albanian army going to take
over the world?”
J e ff B e wk e s , C E O , Ti m e Wa r n e r 2 0 1 0
Ignore at your own peril…
8. Technology is driving fundamental change
T h e c a t a l y s t
Then
Now
Mortgages online
Virtual estate agents
House sharing
Liquidity
Connected homes
Rental listings
Sales listings
Location based
insurance
AirBnB / ZipCar / Uber
Autonomous cars
Tickets
Holidays
Grocery delivery
Customized nutrition
Rent a wardrobe
Automated stores
Loyalty programs
PoS
Netflix
Restaurant to home
Social
Sensory interfaces
Reviews
Mail order books
tickets
Doctors on the phone
Editing DNA
Digital trials via
wearables
Electronic records
Workflows
Devices
Housing Transport Food Entertainment Health
9. … and the way a customer interacts and
purchases changes the very nature of
what they buy
Behavior is changing rapidly and constantly…
10. How customers are evolving
T h e c a t a l y s t
The fuel of
expectation
Convenience,
immediacy, value,
meaning, relevance,
authenticity, social
connection…
Zero effort
behavior
Make it effortless to
‘mesh’ and transact
across the user
journey
Integrated with
senses
Make interaction with
products and services
native to my senses
Understand ‘me’
as I am
Think ‘infinite’
segments and
triggers: based on
who I am, what I like
and what I do
Keep my trust
Treat me with respect,
be authentic, maintain
my trust
13. “After several hours and 2 visits, I was nowhere
closer to getting a joint account with my wife. I
could open one on Monzo in minutes. I closed
my bank account next week”
The voice of the customer…
C u s t o m e r a n e c d o t e
14. Not quite
David and
Goliath…
Focus on customer needs
Abstracting customer need to its simplest and clearest form and
making it your business’s raison d'être
… but a shift in what it takes to
win repeatedly
The ‘easy bits’ are done
The days of information banking, low-touch businesses
are over. Instead, winning now involves going deeper
and creating richer, high-touch experiences which can be
delivered in real time
Move quickly & harness new capabilities
Utilize high impact capabilities to give your business the
edge against competitors
15. The impact (and risks) are clear
44
55
65
72
0 20 40 60 80
Difficult purchase/check-out
No personalised comms
Needs not anticipated
No Easy to use mobile
experience
Percentage of Consumers Who Are Extremely or Somewhat
Likely to Switch Brands if:
Source: Oracle, Salesforce, EuroMonitor
A LSE study showed an average NPS
increase of 7% correlates with a 1% growth
in revenue.
McKinsey states that brands that improve the
customer journey can see revenues increase
as much as 10-15% whilst lowering costs
by 15-20%.
16. Levers to accelerate impact
L e a r n i n g
Data
Ability to build and analyse multi-
dimensional customer datasets
Decentralized rails
Ability to bypass existing
‘guardians’ and their rules
Machine Learning
Tools to make sense, build pace
and enhance customer
experience
Business models
New business models to cater for
rapidly evolving product-market
mix
17. Levers to accelerate impact
L e a r n i n g : t o d a y ’ s a r e a o f f o c u s
Data
Ability to build and analyse multi-
dimensional customer datasets
Decentralized rails
Ability to bypass ‘guardians’
and their rules
Machine Learning
Tools to make sense, build pace
and enhance customer
experience
Business model
New business models to cater for
rapidly evolving product-market
mix
19. Machine Learning: level set
G u i d i n g i d e a s
Or perhaps…?
• Automation
• Real time analysis
• Pattern recognition
• Prediction
• Human emulation
(Note: not Deus Ex
Machina)
20. How ML is being applied in a customer-first context
G u i d i n g i d e a s
Applications of ML
Customer
Service
Fraud
prevention
Personal-
ization
Service
enhancement
Logistics
Insight
- Infinite segmentation
- Predictive models to
forecast need
- Virtual assistants
- Chatbots
- Appending metadata
- Reducing customer
friction
- Detecting sophisticated
fraud patterns
- Customer protection
- Demand forecasting
models
- Hyper
personalization of
content and
journeys
Real time
outreach
- Real time ‘nudges’
Marketing
22. 1. Customer First: Start with the right mindset
G u i d i n g i d e a s
• Integral to the mission statement of your
business
• Deeply integrated into how the business
operates and grows
• A way to guide your decisions
• Stretch time horizons: short to long term
• Starting point for culture change: align all
aspects of the organisation
• Judgment exercise: need to get balance right
What it should be:
• A mechanism to ignore or deprioritise other
parts of your business: the entire business
aligns to deliver for the customer
• A reason to make lots of changes to plan in
response to every piece of feedback you
receive: seek feedback but develop metrics
to drive actions over the long term
• An excuse to think only about what’s around
the corner and deprioritize GSD: focus on
execution and keep iterating
What it shouldn’t become:
23. 2. Leadership aligns, asks the right questions and sets a clear
strategy
G u i d i n g i d e a s
• Focus on the problem from the customer’s
perspective and revisit often(e.g. difficult to make
decisions at different point in user journey due to lack
of information)
• Translate that into a business problem that needs to
be solve and set goals
• Ask: “do we really need ML for this”?
• Understand what is being ‘learned’ and how ML will be
applied (automation, personalisation, etc.) and how
the solution will scale
• Educate and align: take everyone along (i.e. does the
Head of Sales get it?). Cascade down to middle
management
• Solutions will cut across siloes so beware frictions
(“my turf”). Board level alignment helps
• If you are using 3P providers: get past the ‘guff’ and
focus on use cases / impact
Board/CEO:
“What should my AI/ML strategy be?”
Director of Marketing:
“everyone is doing it and ML
makes everything so exciting!”
Director of Product/ Engineering:
“… you don’t have any idea what it actually
is. My focus is keeping the lights on!”
24. 3. Integrate deeply in the core business
G u i d i n g i d e a s
• Entrench deep in the business where expertise and knowledge lies (external centres of
excellence come with risks of reduced pace and/or difficulty in embedding outputs)
• Integrate into the core (or ‘every-day’) activities (planning, marketing, customer service)
• Empower and skill team to develop, test and drive use cases (brings down cost, increases
speed and impact)
• Measure?
• % of Pan BU/organisational processes that are ML-based?
• % of IT budget dedicated to analytics / ML?
25. 4. Execute: set up the right ways of working
G u i d i n g i d e a s
Prioritise
Analyse
Implement
Review
Customer
need
Measure
Hypothesis
Experiment
Specific ideas based on
judgment and observation
The right metrics to
a) measure behavior
b) criterion for
success
Prioritising based on
impact, resourcing
and complexity
Executing scale trials
PMs stay close to channels
Dispassionately
review outcomes,
iterate / kill as
necessary
Implement in the real
world
Measure impact on
decisions and
business, iterate
• Champion evidence and get
feedback
• Be hypothesis and metrics led to
maintain focus
• PMs/marketers maintain links
with sales and customer support
channels
• Celebrate successes and failure;
progress with each iteration – hit
or miss
• Measure?
• Recall (%) and accuracy (%) rates
achieved
• Input: % of customers who took
recommended action
• Output: Post X weeks attributed
impact?
26. 5. Refine your
data strategy
G u i d i n g i d e a s
Manage data risk but
democratise; allow
downstream flexibility
Users
Set up the right governance
models to for accountability,
maintenance (critical vs. not),
optimise cost
Governance
Create infrastructure that caters
for vast majority of use cases;
highly sophisticated use cases
can be catered for by exception
Infrastructure
Provide tools that
accelerate teams’ ability
to experiment, raise
standard on
methodologies
Tools & methodology
27. 6. Obsess about talent
G u i d i n g i d e a s
Set the right hiring bar
Set a genuine hiring message
Clusters and impact: strong stories do wonders
Key management roles, Chief analytics Officer?
Hire
Make roles integral to team / BU / Company success
Review and create career paths
Rotate via different paths to build expertise
Develop
Retain focus on existing talent – many want to learn
Strong motivating mechanism
Communities can beat classrooms!
Skill
Resource is scarce and valuable
Empower and enable
Critical mass is key
Retain • Analytical talent is
expensive, hard to
recruit, difficult to
motivate/ integrate into
the business and even
harder to retain
• Needs a mind-set shift
• Measure?
• Analytical staff / 100
FTEs?
• % of Level 1 trained
employees per team?
• % of Level 2 trained
employees per team?
• % of goals set that have
a material analytical
component?
28. 7. Organise for skill
G u i d i n g i d e a s
• Cross functional teams are key to breaking siloes and driving scale projects across the business
• Regardless of model, staff teams for skill and a rapid test-iterate way of working
• Establish accountability for decision making
• Anchor program outputs to key metrics (e.g. sign-on score)
Program
Manager
Head of
Digital
Director
Digital
CEO
SDE
Head of
Engg
CTO
Marketing
Manager
Head of
Marketing
Director
Marketing
UX
Designer
Data
Scientist
BI Engineer
Head of
Reporting
Director
Finance
Program
Manager
Data
Scientist
BI Engineer
Marketing
Manager
Head of
Marketing
Director
Marketing
CEO
Understand need across channels
Surface days are gone – you can no longer be Yelp – you need to be a Deliveroo
Adopt the right technology and business model
Don’t just ‘think’ about the customer … make it the driving force of every part of your business