5. 5
Machine Learning … Generally Speaking
Used with structured data
sets
Good for unsupervised,
supervised, and semi
supervised
6. 6
Deep Learning … Generally Speaking
Used with unstructured
data sets
Good for supervised, and
semi supervised data
7. Conversion Rates
Tracking conversion rates
to train models to predict
conversion rates on
unseen data.
Improve CRM
Opportunities
Identify the best
combination of attributes
to suggest process
improvements.
Predict Churn
Predict churn rates for
new customers based on
historical patterns.
Market Segments
Cluster similar personas
into market segments to
create more targeted
marketing campaigns.
7
Use Cases with Machine Learning
8. Image
Recognition
Assign attribution to
offline marketing
campaigns.
Measure Intent
Use sentiment analysis to
gage intent for a
potential customer.
Translations
Reach out to potential
customers that are not
fluent in your native
language.
Voice Patterns
Identify the most
common voice patterns
for you most successful
sales reps.
8
Use Cases with Deep Learning
10. How do we setup the
organization to implement ML
and DL?
11. Hacking Skills
Programming, data
munging
Domain Level
Expertise
The best data scientists
are those that
understand the problems
they are try. They
understand the Sales
and Marketing domain!
11
The Data Science Persona
Math and Stats
Mathematical skills,
mostly involved with
statistics, algebra, and
some calc.
The Data
Scientist
The ideal data scientist
has skills from all three
domains!
12. 12
Low Hanging Fruit - Quick Wins
Improve
performance of a
Landing Page
Use a chat bot
and predict high
value customers
Measure
sentiment on
social media
Predictive
lead scoring
13. We are offering too many
discounts
The firm is offering blanket discounts to all of
their potential customers to reach end of
quarter goals.
13
Try to Solve a Business Problem
Segment users
Segment users based on similar attributes.
Later stages can focus on micro
segmentation.
Experiment with A/B Tests
It’s only a science if you can test it. Test
various campaigns with various segments
and test them. Encourage challenger models.
14. 14
Key Takeaways
Machine
Learning and
Deep
Learning is a
process
You can’t get around the data
munging, for now, anyway.
Deep Learning is used mostly for
supervised learning problems
Automating the ML and DL
pipelines are important
Data science is a team effort
A.I. doesn’t exist yet. But it’s less
of a mouth full.
16. 16
B2B Accout Propensity Scores
The combination
of ideal Fit /
Intent Data is
unique
Account ‘Fit’ Information
Fit Information includes
Firmographic, Demographic, and
Technographic data
Behavior = Intent
Behavior on channels that aren’t just
public web sites
17. 17
How Cup of Data Delivers Leads
Training
Data
Testing
Data
ML/DL
Models
Cross
Validatio
n
Cup of
Data
Optimizat
ion
Engine
Conversion
Rates
New
Configurations
API
Sales
and
Martech
Apps
Data with
conversion
results
InsideView
Fit Data
SQL DB
Deliverable to Client
= (Account/Lead +
Context + Account
Propensity Score
(APS))
Behavior
Data
CoD API
Gatewa
y
Third party integration
services: Zapier, Mulesoft,
Slack, etc.
Graph
DB
CoD
App
Services