Abstract:- Discussion of various applied cases of artificial intelligence solutions stemming from our venture work across verticals including consumer, financial and media.
1. Disrupting Corporates with AI
Aug 2017
Sudipto Das
Director of Engineering
BCG Digital Ventures
Hicham Mhanna
VP of Engineering
BCG Digital Ventures
2. 2
DV Centers
Manhattan Beach
Berlin
Sydney
London
Tokyo
DV Labs
Silicon Valley
San Francisco
Seattle
Mexico City
Shanghai
Our Mission
To invent, launch, and scale game-changing businesses with the worldâs most
important corporations and partners
A Unique Combination
BCGDV Innovation | Incubation | Commercialization
BCG Strategy | Transformation
B Capital Corporate Ventures | Market Investment
500+ DVers globally
6 centers & 5 labs
60+ corporate partners
150+ ventures
20+ active incubations
Our multi-disciplinary startup teams
Venture Architects| Product Managers | Strategic Designers
Experience Designers | Engineers | Data Scientists | Growth Hackers
3. 3
&
When we find an
untapped source
of opportunity, we
When we find an
existing venture with
high potential, we
new products
and businesses
in IP, talent and
market opportunities
InvestInvent
We are a Venture Capital firm, and much more
4. 4
Case Study - Personalization
Goal
Become the most
personalized brand in the
world
Starting point
Segmented, not 1:1
targeting
Highly manual
operations
Simple
gamification
Single channel
personalization
5. 5
âą Simplifies with segmentation
âą Assumes the same journey
âą Interactions on different channels are independent
Traditional Approach
6. 6
âą Library of dynamic experiences
âą Individualised experience assignment triggered by
event
âą The system reacts, predicts and learns
âą unique journeys are created to maximise response
Example
Takt detects that John Smith (a latte drinking Seattleite) enters a store in
Vancouver, Canada for the first time.
Based on this event data, an individualised experience is instantly
activated for John.
He receives a recommendation to try the New maple Latte (a new
menu item exclusive to Canada that is known to be a hit with Latte
lovers visiting from the US)
TAKT approach
7. 7
Email example
xx million
2-digit variants
6-digit variants
Segmentation Individualization
xx million
6-digit variants
2-digit variants
No. People :
No. Variants :
People/variant:
Multiple machine learning models individualize the message for
content as well as the point on the economic efficiency curve.
The sole reason for a cluster of >1 is that some people simply
like the same thing.
Dynamic construction of the message
Real-time tracking of the progress for the experience
8. 8
Engineering Footprint
Ingestion Services Platform Services Output Services
OPERATIONAL SERVICES
Domain 3rd Party APIs Channel
Transaction Data
Email Events
Customer
Information
Product Data
. . .
. . .
Weather
Data
. . .
. . .
. . .
. . .
. . .
Recommendation
API
Reward
API
. . .
. . .
Email
Mobile
. . .
. . .
. . .
. . .
. . .
. . .
. . .
. . .
Workflow
tracking
Audience
Partitioning
Preference
Models
Product
Hierarchy
Process
Orchestrator
Experimentation
Context
Models
Workflow
Configuration
Business Rules
Workflow
Creation
Federated ML
Services
Asset
Management
Measurement
Optimization
Models
Other
Individualization Services
Workflow Services
Content Services
Template
Management
Propensity
Models
Collaborative
Filter
Churn
Model
Recommendation
Model
9. 9
What is Taktâs customer engagement platform
- Ingesting 100M+ events a day
- Advanced Machine Learning: multiple
dimensions individualize touchpoints for
20M customers
- Real-time ââcustomer stateââ tracking
- Integrated email and mobile channels
- Scalable, API-based cloud architecture
- Delivering:
- 60M individualized emails per month
- 230M individualized mobile sessions per
month
Takt today -- An advantaged software company
Takt Fakts
- Stand-alone Enterprise SAAS
company
- 67 people; 80% engineering, data
science and product; engineers
from Google, Amazon, Salesforce
- $30M series A funding
- Long-term contract with Starbucks
10. 10
Impact delivered
HOWARD SCHULTZ, Chairman
"Our new one-to-one personalized marketing capability...will prove to be a retail industry
game changerâ
KEVIN JOHNSON, CEO
âThe personalized Star Dashes and suggested selling are examples of how we are using
personalization to provide a relevant customer experience and to increase engagement...we
saw 8% growth, the highest growth rate in average spend per active rewards member over a
prior year ever, reflecting both increased ticket and transaction frequency."
Source: Starbucks Investor Day â December, 2016 and April, 2017 earnings call