SlideShare a Scribd company logo
Product
Management
in the era of
Data Science
Mandar Parikh
@mandarparik
h
VP Product & Engineering
@ Entytle
Ex-
Aviso
Zuora
Salesforce
Today’s Webinar
• What does it take to build killer
products in the “AI-first” era?
• What makes for a great Data
Science-driven product and how do
great Product Managers leverage
Data Science to drive value for
customers?
• Find out how to avoid the pitfalls of
hype-chasing Data Science tactics.
• Learn how to work with Data Science
and Engineering to build a compelling
product and solve real problems.
Product
Data
Science
Engineering
Agenda
1. What is Data Science? What are Data Science-enabled
Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DS Teams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
1. What is Data Science? What are Data Science-enabled
Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DS Teams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
What is “Data Science”?
• Not really science: not falsifiable, often untested, no explanatory
power
• An approach that analyzes large amounts of data (where large
is relative) and extracts patterns and insights and makes
predictions from it, to impact some business driver
• New term for statistical techniques that have existed for
decades
• Data Science is at the “core” of the offering; differentiating IP
• Tons of examples
• LinkedIn
• Uber
• Salesforce
• I won’t talk about other applications of AI including NLP, Image
Recognition et al.
1
What are “Data Science-enabled Products”?
What are “Data Science-enabled
Products”?
LinkedIn
• Early adopters of Data Science
• Graph search
• Skills
• Reputation
• Recruiting
• “People You May Know”
2
What are “Data Science-enabled
Products”?
Uber
• Driver Positioning
• Ride Demand Forecasting
• Surge Pricing
• Estimating ETA
3
• Started as a Contact
Manager, no Data
Science
• Salesforce Einstein
• “Provides Insights to
Users”
4What are “Data Science-enabled
Products”?
Salesforce
1. What is Data Science? What are Data Science-enabled
Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DS Teams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
Data Science Basics
• Features & Feature Selection
• ”Art of Data Science is to figure out which features to use”
• Unsupervised Learning
• “What patterns exist in the data?”
• Supervised Learning
• “Make predictions using existing patterns in the data”
• Popular algorithms: Random Forest, Gradient Boosted
Machines, Kernel Regression, Logistic Regression, Nearest
Neighbors, Clustering
• Deep Learning
A Word On Machine Learning
• DS, ML and AI used interchangeably
• I am not here to settle the debate, but I think of it as:
• DS model extracts patterns and makes predictions
• ML automatically calibrates the model and improves predictions over
time
Different
Types of
Analyses
Descriptive:
What happened?
01
Predictive:
What will
happen?
02
Prescriptive:
What should I do
(to achieve some
outcome)?
03
DESCRIPTIVE ANALYTICS
PREDICTIVE
ANALYTICS
Prescriptive Analytics
Price Optimization
1
Expert Systems /
Decision Support
Systems
2
What-if Systems
3
Accuracy v. Recall v. Precision v…..
Business Metrics
N=170 Predicted: YES Predicted: NO
Actual: YES 70 5
Actual: NO 45 50
Accuracy: How many times were you right? (70+50)/(70+50+5+45) = 70%
Recall: Of all the positives, how many times were you right? 70/(70+5) = 93%
Precision: How many times did you correctly predict YES? 70/(70+45) = 61%
But none are Business Metrics!
1. What is Data Science? What are Data Science-enabled
Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DS Teams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
Product Management Fundamentals Don’t Change
• Customer Empathy
• Data-driven
Salesforce Lead Scoring
Tell me why a lead is an 83
v.
#1
Cleanliness is Next to Godliness
• The data you get will not be clean – guaranteed
• Data cleansing is the single best thing you can do for your
data, beating incremental improvements you’ll get with a
”better” model
• Cleaning data is HARD! Don’t underestimate the effort involved.
#2
• Don’t over think it, and don’t let
your Data Scientists tell you
otherwise
• Start simple; you might be
surprised how far you can get
• Don’t get stuck in Analysis –
Paralysis; be skeptical of
incremental improvements
#3The Best Model is Often
the Simplest One
Get the AI Out of the Way
Opportunity Score
Opportunity 42929 48%
Opportunity 23427 68%
Opportunity 55620 23%
Opportunity Score
Opportunity 42929 Medium
Opportunity 23427 High
Opportunity 55620 Low
#4
Make it Actionable
Opportunity Score Why?
Opportunity 42929 Medium Time-in-stage,
Meeting/email sentiment
Opportunity 23427 High High level of activity,
External factors
Opportunity 55620 Low Close date pushed
AKA Interpretability
#5
Your Users Are NOT Data Scientists
No need to
show the
confidence
level or
margin of
error
Users like plain-speak
explanations.
#6
Design is AI’s Best Friend
Cadence
is key
Progressive Disclosure
Drill-down
Flexible &
uncluttered
Trend arrows
are great
Name the segments
#7
• Less is more
• Show key insights, skip the details (but be
able to drill-in)
• Articulate the insights, don’t keep the user
guessing
• Anomaly Detection is powerful at scale and is
great for ”management by exception”
• If you’re using Clustering, it’s often insightful
to expose the output to managers
• Managers love Prescriptive Analytics;
consider What-If modeling
http://www.tatvic.com/blog/wp-
content/uploads/2017/01/fetured.jpg
Prescriptive Analytics
Anomaly Detection
#8
Managers’ Deserve Less!
Tools
■ BITools LikeTableau, Domo
■ Excel, R, Jupyter for quick
prototyping
■ D3.js – very popular for building
rich, interactive visualizations
Techniques
■ “TheVisual Display of Quantitative
Information”, by EdwardTufte
■ …
Don’t Visualize Data:
Tell Stories
#9
Be The QA You Want To See
KPIs & Metrics
01
Backtesting
02
#10
Backtesting:
The Basic Idea
1. Take a big slice of valid data
2.Split into training and testing data
sets
a. Training data set should be several
multiples of “business cycles”
b. Testing data set should probably
be a few “business cycles”
3. Train the model on the training
data set
4.Run the model on the testing
data set and generate predictions
over the testing period
5. Compare actuals to predictions in
the testing period
Backtesting: Motivations
• Basis for comparison for new stuff from Data Science; baseline for
measuring marginal value added
• Regression test for Engineering
• Quantifiable quality test for Product & Customer Success
• Aid in Sales and Marketing
1. What is Data Science?What are Data Science-enabled Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DSTeams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
All Big Data Is Not Equal
• Particularly in Enterprise systems, the amount of data is not “Big”
• Can make useful predictions or provide useful insights on relatively
small amount of data
• E.g.Years worth of Opportunity Data <<A month’s worth of Censor Data
• More data does not necessarily mean better predictions
• E.g.Very old data may be “out-of-date” and may impact accuracy negatively
• You don’t need Data Science
ALL the time
• Don’t shy away from a “simple"
solution just because there is
no DS involved
Data Science is NOT a Silver Bullet
1. What is Data Science? What are Data Science-enabled
Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DS Teams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
Engaging with DS Teams
• Integral part of the Product & Engineering Team
• Much of Data Science is (also) Software Engineering
• Should be part of the “normal” product cadence and processes
• Bring in Data Science early into Customer conversations (a la
Design)
• Drive use-case-based initiatives with Data Science
1
Engaging with DS Teams
• Help DS understand and define the problem
• Help with the art, leave the science to DS
• Help DS understand the context, assist in feature-selection
• DS picks the algorithms
• Much like API design, agree on the contract: inputs and outputs
• Don’t forget the UX part of Data Science
• Reinforce Business Metrics
2
1. What is Data Science? What are Data Science-enabled
Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DS Teams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
• As the spokesperson for your product, learn to speak Data
Science
• You will have to grapple with:
• When to mention DS and when not to?
• When you speak about DS, how much do you give-away?
• How do you explain algorithms to the lay user?
• Understand your audience, understand the context
1
Product Marketing with Data Science
Context
• Consumer App: “Powered by sophisticated AI” works
• Business App: “Powered by sophisticated AI” could back-fire
with users who have been following a process for decades
2
Product Marketing with Data Science
What to say & How to say it
• Check with your DS team and Legal/IP on what NOT to say
• General approach to solving the problem is probably okay
• IP is usually in the end-to-end solution, not the choice of model
• Stick with simple explanations that your grandmother could
understand
• Be prepared to back up your claims (proof-points)
3
Product Marketing with Data Science
1. What is Data Science? What are Data Science-enabled
Products?
2. A Mini-Primer on Data Science
3. Designing Products with Data Science: Strategies for Success
4. Data Science Pitfalls
5. Engaging with DS Teams
6. Product Marketing & Data Science
7. Data Science Essentials for Product Managers
Data Science Essentials for Product Managers
• A background in basic statistics is an absolute must
• Pick up a book with a “high-level” intro to Data Science; my favorites are:
• Numsense! Data Science for the Layman: No Math Added; by Annalyn Ng, Kenneth
Soo
• Data Science for Dummies, Lillian Pierson
• An “Intro to Data Science” online class could be useful; here’s one: Intro to
Data Science (Udemy)
• Unlike Balsamiq or Sketch for UX Design, there are no “starter” toolkits for
Data Science-related prototyping
• Understanding Jupyter (or similar) output (metrics, curves, distributions) is
a good skill to develop
• Excel or analytic tools like Looker are great arsenal to have
• Ability to collaborate closely with your Data Scientists is key
Thank You
@mandarparikh

More Related Content

What's hot

Agile Data Science
Agile Data ScienceAgile Data Science
Agile Data Science
Volodymyr Kazantsev
 
UX Field Research Basics
 UX Field Research Basics  UX Field Research Basics
UX Field Research Basics
David Farkas
 
From SCRUM to LEAN
From SCRUM to LEANFrom SCRUM to LEAN
From SCRUM to LEAN
Vlad Mysla
 
High-Impact Analytics for Mortals
High-Impact Analytics for MortalsHigh-Impact Analytics for Mortals
High-Impact Analytics for Mortals
T. Scott Clendaniel
 
Data-Driven Requirements for User-Stories on JustAnswer
Data-Driven Requirements for User-Stories on JustAnswerData-Driven Requirements for User-Stories on JustAnswer
Data-Driven Requirements for User-Stories on JustAnswer
Vlad Mysla
 
UX Field Research Basics Chicago Camp 2017
UX Field Research Basics Chicago Camp 2017UX Field Research Basics Chicago Camp 2017
UX Field Research Basics Chicago Camp 2017
David Farkas
 
H2O World - Advanced Analytics at Macys.com - Daqing Zhao
H2O World - Advanced Analytics at Macys.com - Daqing ZhaoH2O World - Advanced Analytics at Macys.com - Daqing Zhao
H2O World - Advanced Analytics at Macys.com - Daqing Zhao
Sri Ambati
 
Machine Learning/ Data Science: Boosting Predictive Analytics Model Performance
Machine Learning/ Data Science: Boosting Predictive Analytics Model PerformanceMachine Learning/ Data Science: Boosting Predictive Analytics Model Performance
Machine Learning/ Data Science: Boosting Predictive Analytics Model Performance
T. Scott Clendaniel
 
Leveraged Analytics at Scale
Leveraged Analytics at ScaleLeveraged Analytics at Scale
Leveraged Analytics at Scale
Domino Data Lab
 
Radical Analytics - SimpliLearn Webinar
Radical Analytics - SimpliLearn WebinarRadical Analytics - SimpliLearn Webinar
Radical Analytics - SimpliLearn Webinar
Stéphane Hamel
 
JDO 2019: Data Science for Developers - Matthew Renze
JDO 2019: Data Science for Developers -  Matthew RenzeJDO 2019: Data Science for Developers -  Matthew Renze
JDO 2019: Data Science for Developers - Matthew Renze
PROIDEA
 
Anatomy of a data science project
Anatomy of a data science projectAnatomy of a data science project
Anatomy of a data science project
Adam Sroka
 
Intranet Review Toolkit
Intranet Review ToolkitIntranet Review Toolkit
Intranet Review Toolkit
Elizabeth Lupfer
 
1645 track 1 bress_using his laptop
1645 track 1 bress_using his laptop1645 track 1 bress_using his laptop
1645 track 1 bress_using his laptop
Rising Media, Inc.
 
Andreas weigend
Andreas weigendAndreas weigend
Andreas weigend
BigDataExpo
 
5. Workshop Responsible Data Science - Discussion on Transparency in data sci...
5. Workshop Responsible Data Science - Discussion on Transparency in data sci...5. Workshop Responsible Data Science - Discussion on Transparency in data sci...
5. Workshop Responsible Data Science - Discussion on Transparency in data sci...
Jheronimus Academy of Data Science
 
Operationalizing Machine Learning
Operationalizing Machine LearningOperationalizing Machine Learning
Operationalizing Machine Learning
AgileThought
 
Notilyze SAS
Notilyze SASNotilyze SAS
Notilyze SAS
BigDataExpo
 
Make good products great with data and analytics
Make good products great with data and analyticsMake good products great with data and analytics
Make good products great with data and analytics
David Mathias
 
10 Tips for women to build a career in data science
10 Tips for women to build a career in data science10 Tips for women to build a career in data science
10 Tips for women to build a career in data science
Carol Hargreaves
 

What's hot (20)

Agile Data Science
Agile Data ScienceAgile Data Science
Agile Data Science
 
UX Field Research Basics
 UX Field Research Basics  UX Field Research Basics
UX Field Research Basics
 
From SCRUM to LEAN
From SCRUM to LEANFrom SCRUM to LEAN
From SCRUM to LEAN
 
High-Impact Analytics for Mortals
High-Impact Analytics for MortalsHigh-Impact Analytics for Mortals
High-Impact Analytics for Mortals
 
Data-Driven Requirements for User-Stories on JustAnswer
Data-Driven Requirements for User-Stories on JustAnswerData-Driven Requirements for User-Stories on JustAnswer
Data-Driven Requirements for User-Stories on JustAnswer
 
UX Field Research Basics Chicago Camp 2017
UX Field Research Basics Chicago Camp 2017UX Field Research Basics Chicago Camp 2017
UX Field Research Basics Chicago Camp 2017
 
H2O World - Advanced Analytics at Macys.com - Daqing Zhao
H2O World - Advanced Analytics at Macys.com - Daqing ZhaoH2O World - Advanced Analytics at Macys.com - Daqing Zhao
H2O World - Advanced Analytics at Macys.com - Daqing Zhao
 
Machine Learning/ Data Science: Boosting Predictive Analytics Model Performance
Machine Learning/ Data Science: Boosting Predictive Analytics Model PerformanceMachine Learning/ Data Science: Boosting Predictive Analytics Model Performance
Machine Learning/ Data Science: Boosting Predictive Analytics Model Performance
 
Leveraged Analytics at Scale
Leveraged Analytics at ScaleLeveraged Analytics at Scale
Leveraged Analytics at Scale
 
Radical Analytics - SimpliLearn Webinar
Radical Analytics - SimpliLearn WebinarRadical Analytics - SimpliLearn Webinar
Radical Analytics - SimpliLearn Webinar
 
JDO 2019: Data Science for Developers - Matthew Renze
JDO 2019: Data Science for Developers -  Matthew RenzeJDO 2019: Data Science for Developers -  Matthew Renze
JDO 2019: Data Science for Developers - Matthew Renze
 
Anatomy of a data science project
Anatomy of a data science projectAnatomy of a data science project
Anatomy of a data science project
 
Intranet Review Toolkit
Intranet Review ToolkitIntranet Review Toolkit
Intranet Review Toolkit
 
1645 track 1 bress_using his laptop
1645 track 1 bress_using his laptop1645 track 1 bress_using his laptop
1645 track 1 bress_using his laptop
 
Andreas weigend
Andreas weigendAndreas weigend
Andreas weigend
 
5. Workshop Responsible Data Science - Discussion on Transparency in data sci...
5. Workshop Responsible Data Science - Discussion on Transparency in data sci...5. Workshop Responsible Data Science - Discussion on Transparency in data sci...
5. Workshop Responsible Data Science - Discussion on Transparency in data sci...
 
Operationalizing Machine Learning
Operationalizing Machine LearningOperationalizing Machine Learning
Operationalizing Machine Learning
 
Notilyze SAS
Notilyze SASNotilyze SAS
Notilyze SAS
 
Make good products great with data and analytics
Make good products great with data and analyticsMake good products great with data and analytics
Make good products great with data and analytics
 
10 Tips for women to build a career in data science
10 Tips for women to build a career in data science10 Tips for women to build a career in data science
10 Tips for women to build a career in data science
 

Similar to Product Management in the Era of Data Science

Lean Analytics: How to get more out of your data science team
Lean Analytics: How to get more out of your data science teamLean Analytics: How to get more out of your data science team
Lean Analytics: How to get more out of your data science team
Digital Transformation EXPO Event Series
 
InfoVision_PM101_RPadaki
InfoVision_PM101_RPadakiInfoVision_PM101_RPadaki
InfoVision_PM101_RPadaki
Ravi Padaki
 
Data and data scientists are not equal to money david hoyle
Data and data scientists are not equal to money   david hoyleData and data scientists are not equal to money   david hoyle
Data and data scientists are not equal to money david hoyle
Institute of Contemporary Sciences
 
Product Management 101 for Data and Analytics
Product Management 101 for Data and Analytics Product Management 101 for Data and Analytics
Product Management 101 for Data and Analytics
Ravi Padaki
 
lec1.pdf
lec1.pdflec1.pdf
lec1.pdf
nimmakiran1
 
BEST Practices - Testing & Optimization | Bredan Rendan
BEST Practices - Testing & Optimization | Bredan RendanBEST Practices - Testing & Optimization | Bredan Rendan
BEST Practices - Testing & Optimization | Bredan Rendan
Caleb Whitmore
 
Tableau Conference 2014 Presentation
Tableau Conference 2014 PresentationTableau Conference 2014 Presentation
Tableau Conference 2014 Presentation
krystalstjulien
 
Big Data & Business Strategy
Big Data & Business StrategyBig Data & Business Strategy
Big Data & Business Strategy
Sylvia Ogweng
 
Synergis60: 6 Critical Steps to Implementing Data Managment
Synergis60: 6 Critical Steps to Implementing Data ManagmentSynergis60: 6 Critical Steps to Implementing Data Managment
Synergis60: 6 Critical Steps to Implementing Data Managment
Synergis Engineering Design Solutions
 
Managing Data Science | Lessons from the Field
Managing Data Science | Lessons from the Field Managing Data Science | Lessons from the Field
Managing Data Science | Lessons from the Field
Domino Data Lab
 
What Managers Need to Know about Data Science
What Managers Need to Know about Data ScienceWhat Managers Need to Know about Data Science
What Managers Need to Know about Data Science
Annie Flippo
 
Advanced Project Data Analytics for Improved Project Delivery
Advanced Project Data Analytics for Improved Project DeliveryAdvanced Project Data Analytics for Improved Project Delivery
Advanced Project Data Analytics for Improved Project Delivery
Mark Constable
 
Tableau Drive, A new methodology for scaling your analytic culture
Tableau Drive, A new methodology for scaling your analytic cultureTableau Drive, A new methodology for scaling your analytic culture
Tableau Drive, A new methodology for scaling your analytic culture
Tableau Software
 
10 Tips From A Young Data Scientist
10 Tips From A Young Data Scientist10 Tips From A Young Data Scientist
10 Tips From A Young Data Scientist
Nuno Carneiro
 
data science and business analytics
data science and business analyticsdata science and business analytics
data science and business analytics
sunnypatil1778
 
Data-Driven Organisation
Data-Driven OrganisationData-Driven Organisation
Data-Driven Organisation
Jaakko Särelä
 
Doing Analytics Right - Building the Analytics Environment
Doing Analytics Right - Building the Analytics EnvironmentDoing Analytics Right - Building the Analytics Environment
Doing Analytics Right - Building the Analytics Environment
Tasktop
 
AdTechBLR_HowToMakeDataActionable
AdTechBLR_HowToMakeDataActionableAdTechBLR_HowToMakeDataActionable
AdTechBLR_HowToMakeDataActionable
Ravi Padaki
 
Big data@work
Big data@workBig data@work
Why is your model stuck in the lab? How to move your model from the lab to pr...
Why is your model stuck in the lab? How to move your model from the lab to pr...Why is your model stuck in the lab? How to move your model from the lab to pr...
Why is your model stuck in the lab? How to move your model from the lab to pr...
Data Con LA
 

Similar to Product Management in the Era of Data Science (20)

Lean Analytics: How to get more out of your data science team
Lean Analytics: How to get more out of your data science teamLean Analytics: How to get more out of your data science team
Lean Analytics: How to get more out of your data science team
 
InfoVision_PM101_RPadaki
InfoVision_PM101_RPadakiInfoVision_PM101_RPadaki
InfoVision_PM101_RPadaki
 
Data and data scientists are not equal to money david hoyle
Data and data scientists are not equal to money   david hoyleData and data scientists are not equal to money   david hoyle
Data and data scientists are not equal to money david hoyle
 
Product Management 101 for Data and Analytics
Product Management 101 for Data and Analytics Product Management 101 for Data and Analytics
Product Management 101 for Data and Analytics
 
lec1.pdf
lec1.pdflec1.pdf
lec1.pdf
 
BEST Practices - Testing & Optimization | Bredan Rendan
BEST Practices - Testing & Optimization | Bredan RendanBEST Practices - Testing & Optimization | Bredan Rendan
BEST Practices - Testing & Optimization | Bredan Rendan
 
Tableau Conference 2014 Presentation
Tableau Conference 2014 PresentationTableau Conference 2014 Presentation
Tableau Conference 2014 Presentation
 
Big Data & Business Strategy
Big Data & Business StrategyBig Data & Business Strategy
Big Data & Business Strategy
 
Synergis60: 6 Critical Steps to Implementing Data Managment
Synergis60: 6 Critical Steps to Implementing Data ManagmentSynergis60: 6 Critical Steps to Implementing Data Managment
Synergis60: 6 Critical Steps to Implementing Data Managment
 
Managing Data Science | Lessons from the Field
Managing Data Science | Lessons from the Field Managing Data Science | Lessons from the Field
Managing Data Science | Lessons from the Field
 
What Managers Need to Know about Data Science
What Managers Need to Know about Data ScienceWhat Managers Need to Know about Data Science
What Managers Need to Know about Data Science
 
Advanced Project Data Analytics for Improved Project Delivery
Advanced Project Data Analytics for Improved Project DeliveryAdvanced Project Data Analytics for Improved Project Delivery
Advanced Project Data Analytics for Improved Project Delivery
 
Tableau Drive, A new methodology for scaling your analytic culture
Tableau Drive, A new methodology for scaling your analytic cultureTableau Drive, A new methodology for scaling your analytic culture
Tableau Drive, A new methodology for scaling your analytic culture
 
10 Tips From A Young Data Scientist
10 Tips From A Young Data Scientist10 Tips From A Young Data Scientist
10 Tips From A Young Data Scientist
 
data science and business analytics
data science and business analyticsdata science and business analytics
data science and business analytics
 
Data-Driven Organisation
Data-Driven OrganisationData-Driven Organisation
Data-Driven Organisation
 
Doing Analytics Right - Building the Analytics Environment
Doing Analytics Right - Building the Analytics EnvironmentDoing Analytics Right - Building the Analytics Environment
Doing Analytics Right - Building the Analytics Environment
 
AdTechBLR_HowToMakeDataActionable
AdTechBLR_HowToMakeDataActionableAdTechBLR_HowToMakeDataActionable
AdTechBLR_HowToMakeDataActionable
 
Big data@work
Big data@workBig data@work
Big data@work
 
Why is your model stuck in the lab? How to move your model from the lab to pr...
Why is your model stuck in the lab? How to move your model from the lab to pr...Why is your model stuck in the lab? How to move your model from the lab to pr...
Why is your model stuck in the lab? How to move your model from the lab to pr...
 

Recently uploaded

Building Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and MilvusBuilding Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and Milvus
Zilliz
 
Video Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the FutureVideo Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the Future
Alpen-Adria-Universität
 
Ocean lotus Threat actors project by John Sitima 2024 (1).pptx
Ocean lotus Threat actors project by John Sitima 2024 (1).pptxOcean lotus Threat actors project by John Sitima 2024 (1).pptx
Ocean lotus Threat actors project by John Sitima 2024 (1).pptx
SitimaJohn
 
“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”
Claudio Di Ciccio
 
Taking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdfTaking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdf
ssuserfac0301
 
TrustArc Webinar - 2024 Global Privacy Survey
TrustArc Webinar - 2024 Global Privacy SurveyTrustArc Webinar - 2024 Global Privacy Survey
TrustArc Webinar - 2024 Global Privacy Survey
TrustArc
 
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdfUnlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Malak Abu Hammad
 
Fueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte WebinarFueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte Webinar
Zilliz
 
Serial Arm Control in Real Time Presentation
Serial Arm Control in Real Time PresentationSerial Arm Control in Real Time Presentation
Serial Arm Control in Real Time Presentation
tolgahangng
 
Choosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptxChoosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptx
Brandon Minnick, MBA
 
Driving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success StoryDriving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success Story
Safe Software
 
Best 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERPBest 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERP
Pixlogix Infotech
 
UiPath Test Automation using UiPath Test Suite series, part 6
UiPath Test Automation using UiPath Test Suite series, part 6UiPath Test Automation using UiPath Test Suite series, part 6
UiPath Test Automation using UiPath Test Suite series, part 6
DianaGray10
 
How to use Firebase Data Connect For Flutter
How to use Firebase Data Connect For FlutterHow to use Firebase Data Connect For Flutter
How to use Firebase Data Connect For Flutter
Daiki Mogmet Ito
 
“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...
“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...
“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...
Edge AI and Vision Alliance
 
Programming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup SlidesProgramming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup Slides
Zilliz
 
20240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 202420240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 2024
Matthew Sinclair
 
AI 101: An Introduction to the Basics and Impact of Artificial Intelligence
AI 101: An Introduction to the Basics and Impact of Artificial IntelligenceAI 101: An Introduction to the Basics and Impact of Artificial Intelligence
AI 101: An Introduction to the Basics and Impact of Artificial Intelligence
IndexBug
 
Climate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing DaysClimate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing Days
Kari Kakkonen
 
Mind map of terminologies used in context of Generative AI
Mind map of terminologies used in context of Generative AIMind map of terminologies used in context of Generative AI
Mind map of terminologies used in context of Generative AI
Kumud Singh
 

Recently uploaded (20)

Building Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and MilvusBuilding Production Ready Search Pipelines with Spark and Milvus
Building Production Ready Search Pipelines with Spark and Milvus
 
Video Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the FutureVideo Streaming: Then, Now, and in the Future
Video Streaming: Then, Now, and in the Future
 
Ocean lotus Threat actors project by John Sitima 2024 (1).pptx
Ocean lotus Threat actors project by John Sitima 2024 (1).pptxOcean lotus Threat actors project by John Sitima 2024 (1).pptx
Ocean lotus Threat actors project by John Sitima 2024 (1).pptx
 
“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”“I’m still / I’m still / Chaining from the Block”
“I’m still / I’m still / Chaining from the Block”
 
Taking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdfTaking AI to the Next Level in Manufacturing.pdf
Taking AI to the Next Level in Manufacturing.pdf
 
TrustArc Webinar - 2024 Global Privacy Survey
TrustArc Webinar - 2024 Global Privacy SurveyTrustArc Webinar - 2024 Global Privacy Survey
TrustArc Webinar - 2024 Global Privacy Survey
 
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdfUnlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
Unlock the Future of Search with MongoDB Atlas_ Vector Search Unleashed.pdf
 
Fueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte WebinarFueling AI with Great Data with Airbyte Webinar
Fueling AI with Great Data with Airbyte Webinar
 
Serial Arm Control in Real Time Presentation
Serial Arm Control in Real Time PresentationSerial Arm Control in Real Time Presentation
Serial Arm Control in Real Time Presentation
 
Choosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptxChoosing The Best AWS Service For Your Website + API.pptx
Choosing The Best AWS Service For Your Website + API.pptx
 
Driving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success StoryDriving Business Innovation: Latest Generative AI Advancements & Success Story
Driving Business Innovation: Latest Generative AI Advancements & Success Story
 
Best 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERPBest 20 SEO Techniques To Improve Website Visibility In SERP
Best 20 SEO Techniques To Improve Website Visibility In SERP
 
UiPath Test Automation using UiPath Test Suite series, part 6
UiPath Test Automation using UiPath Test Suite series, part 6UiPath Test Automation using UiPath Test Suite series, part 6
UiPath Test Automation using UiPath Test Suite series, part 6
 
How to use Firebase Data Connect For Flutter
How to use Firebase Data Connect For FlutterHow to use Firebase Data Connect For Flutter
How to use Firebase Data Connect For Flutter
 
“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...
“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...
“Building and Scaling AI Applications with the Nx AI Manager,” a Presentation...
 
Programming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup SlidesProgramming Foundation Models with DSPy - Meetup Slides
Programming Foundation Models with DSPy - Meetup Slides
 
20240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 202420240607 QFM018 Elixir Reading List May 2024
20240607 QFM018 Elixir Reading List May 2024
 
AI 101: An Introduction to the Basics and Impact of Artificial Intelligence
AI 101: An Introduction to the Basics and Impact of Artificial IntelligenceAI 101: An Introduction to the Basics and Impact of Artificial Intelligence
AI 101: An Introduction to the Basics and Impact of Artificial Intelligence
 
Climate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing DaysClimate Impact of Software Testing at Nordic Testing Days
Climate Impact of Software Testing at Nordic Testing Days
 
Mind map of terminologies used in context of Generative AI
Mind map of terminologies used in context of Generative AIMind map of terminologies used in context of Generative AI
Mind map of terminologies used in context of Generative AI
 

Product Management in the Era of Data Science

  • 2. Mandar Parikh @mandarparik h VP Product & Engineering @ Entytle Ex- Aviso Zuora Salesforce
  • 3. Today’s Webinar • What does it take to build killer products in the “AI-first” era? • What makes for a great Data Science-driven product and how do great Product Managers leverage Data Science to drive value for customers? • Find out how to avoid the pitfalls of hype-chasing Data Science tactics. • Learn how to work with Data Science and Engineering to build a compelling product and solve real problems. Product Data Science Engineering
  • 4. Agenda 1. What is Data Science? What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DS Teams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 5. 1. What is Data Science? What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DS Teams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 6. What is “Data Science”? • Not really science: not falsifiable, often untested, no explanatory power • An approach that analyzes large amounts of data (where large is relative) and extracts patterns and insights and makes predictions from it, to impact some business driver • New term for statistical techniques that have existed for decades
  • 7. • Data Science is at the “core” of the offering; differentiating IP • Tons of examples • LinkedIn • Uber • Salesforce • I won’t talk about other applications of AI including NLP, Image Recognition et al. 1 What are “Data Science-enabled Products”?
  • 8. What are “Data Science-enabled Products”? LinkedIn • Early adopters of Data Science • Graph search • Skills • Reputation • Recruiting • “People You May Know” 2
  • 9. What are “Data Science-enabled Products”? Uber • Driver Positioning • Ride Demand Forecasting • Surge Pricing • Estimating ETA 3
  • 10. • Started as a Contact Manager, no Data Science • Salesforce Einstein • “Provides Insights to Users” 4What are “Data Science-enabled Products”? Salesforce
  • 11. 1. What is Data Science? What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DS Teams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 12. Data Science Basics • Features & Feature Selection • ”Art of Data Science is to figure out which features to use” • Unsupervised Learning • “What patterns exist in the data?” • Supervised Learning • “Make predictions using existing patterns in the data” • Popular algorithms: Random Forest, Gradient Boosted Machines, Kernel Regression, Logistic Regression, Nearest Neighbors, Clustering • Deep Learning
  • 13. A Word On Machine Learning • DS, ML and AI used interchangeably • I am not here to settle the debate, but I think of it as: • DS model extracts patterns and makes predictions • ML automatically calibrates the model and improves predictions over time
  • 14. Different Types of Analyses Descriptive: What happened? 01 Predictive: What will happen? 02 Prescriptive: What should I do (to achieve some outcome)? 03
  • 17. Prescriptive Analytics Price Optimization 1 Expert Systems / Decision Support Systems 2 What-if Systems 3
  • 18. Accuracy v. Recall v. Precision v….. Business Metrics N=170 Predicted: YES Predicted: NO Actual: YES 70 5 Actual: NO 45 50 Accuracy: How many times were you right? (70+50)/(70+50+5+45) = 70% Recall: Of all the positives, how many times were you right? 70/(70+5) = 93% Precision: How many times did you correctly predict YES? 70/(70+45) = 61% But none are Business Metrics!
  • 19. 1. What is Data Science? What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DS Teams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 20. Product Management Fundamentals Don’t Change • Customer Empathy • Data-driven Salesforce Lead Scoring Tell me why a lead is an 83 v. #1
  • 21. Cleanliness is Next to Godliness • The data you get will not be clean – guaranteed • Data cleansing is the single best thing you can do for your data, beating incremental improvements you’ll get with a ”better” model • Cleaning data is HARD! Don’t underestimate the effort involved. #2
  • 22. • Don’t over think it, and don’t let your Data Scientists tell you otherwise • Start simple; you might be surprised how far you can get • Don’t get stuck in Analysis – Paralysis; be skeptical of incremental improvements #3The Best Model is Often the Simplest One
  • 23. Get the AI Out of the Way Opportunity Score Opportunity 42929 48% Opportunity 23427 68% Opportunity 55620 23% Opportunity Score Opportunity 42929 Medium Opportunity 23427 High Opportunity 55620 Low #4
  • 24. Make it Actionable Opportunity Score Why? Opportunity 42929 Medium Time-in-stage, Meeting/email sentiment Opportunity 23427 High High level of activity, External factors Opportunity 55620 Low Close date pushed AKA Interpretability #5
  • 25. Your Users Are NOT Data Scientists No need to show the confidence level or margin of error Users like plain-speak explanations. #6
  • 26. Design is AI’s Best Friend Cadence is key Progressive Disclosure Drill-down Flexible & uncluttered Trend arrows are great Name the segments #7
  • 27. • Less is more • Show key insights, skip the details (but be able to drill-in) • Articulate the insights, don’t keep the user guessing • Anomaly Detection is powerful at scale and is great for ”management by exception” • If you’re using Clustering, it’s often insightful to expose the output to managers • Managers love Prescriptive Analytics; consider What-If modeling http://www.tatvic.com/blog/wp- content/uploads/2017/01/fetured.jpg Prescriptive Analytics Anomaly Detection #8 Managers’ Deserve Less!
  • 28. Tools ■ BITools LikeTableau, Domo ■ Excel, R, Jupyter for quick prototyping ■ D3.js – very popular for building rich, interactive visualizations Techniques ■ “TheVisual Display of Quantitative Information”, by EdwardTufte ■ … Don’t Visualize Data: Tell Stories #9
  • 29. Be The QA You Want To See KPIs & Metrics 01 Backtesting 02 #10
  • 30. Backtesting: The Basic Idea 1. Take a big slice of valid data 2.Split into training and testing data sets a. Training data set should be several multiples of “business cycles” b. Testing data set should probably be a few “business cycles” 3. Train the model on the training data set 4.Run the model on the testing data set and generate predictions over the testing period 5. Compare actuals to predictions in the testing period
  • 31. Backtesting: Motivations • Basis for comparison for new stuff from Data Science; baseline for measuring marginal value added • Regression test for Engineering • Quantifiable quality test for Product & Customer Success • Aid in Sales and Marketing
  • 32. 1. What is Data Science?What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DSTeams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 33. All Big Data Is Not Equal • Particularly in Enterprise systems, the amount of data is not “Big” • Can make useful predictions or provide useful insights on relatively small amount of data • E.g.Years worth of Opportunity Data <<A month’s worth of Censor Data • More data does not necessarily mean better predictions • E.g.Very old data may be “out-of-date” and may impact accuracy negatively
  • 34. • You don’t need Data Science ALL the time • Don’t shy away from a “simple" solution just because there is no DS involved Data Science is NOT a Silver Bullet
  • 35. 1. What is Data Science? What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DS Teams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 36. Engaging with DS Teams • Integral part of the Product & Engineering Team • Much of Data Science is (also) Software Engineering • Should be part of the “normal” product cadence and processes • Bring in Data Science early into Customer conversations (a la Design) • Drive use-case-based initiatives with Data Science 1
  • 37. Engaging with DS Teams • Help DS understand and define the problem • Help with the art, leave the science to DS • Help DS understand the context, assist in feature-selection • DS picks the algorithms • Much like API design, agree on the contract: inputs and outputs • Don’t forget the UX part of Data Science • Reinforce Business Metrics 2
  • 38. 1. What is Data Science? What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DS Teams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 39. • As the spokesperson for your product, learn to speak Data Science • You will have to grapple with: • When to mention DS and when not to? • When you speak about DS, how much do you give-away? • How do you explain algorithms to the lay user? • Understand your audience, understand the context 1 Product Marketing with Data Science
  • 40. Context • Consumer App: “Powered by sophisticated AI” works • Business App: “Powered by sophisticated AI” could back-fire with users who have been following a process for decades 2 Product Marketing with Data Science
  • 41. What to say & How to say it • Check with your DS team and Legal/IP on what NOT to say • General approach to solving the problem is probably okay • IP is usually in the end-to-end solution, not the choice of model • Stick with simple explanations that your grandmother could understand • Be prepared to back up your claims (proof-points) 3 Product Marketing with Data Science
  • 42. 1. What is Data Science? What are Data Science-enabled Products? 2. A Mini-Primer on Data Science 3. Designing Products with Data Science: Strategies for Success 4. Data Science Pitfalls 5. Engaging with DS Teams 6. Product Marketing & Data Science 7. Data Science Essentials for Product Managers
  • 43. Data Science Essentials for Product Managers • A background in basic statistics is an absolute must • Pick up a book with a “high-level” intro to Data Science; my favorites are: • Numsense! Data Science for the Layman: No Math Added; by Annalyn Ng, Kenneth Soo • Data Science for Dummies, Lillian Pierson • An “Intro to Data Science” online class could be useful; here’s one: Intro to Data Science (Udemy) • Unlike Balsamiq or Sketch for UX Design, there are no “starter” toolkits for Data Science-related prototyping • Understanding Jupyter (or similar) output (metrics, curves, distributions) is a good skill to develop • Excel or analytic tools like Looker are great arsenal to have • Ability to collaborate closely with your Data Scientists is key