data @ nflx (tableau conference 2014 presentation)

39
DATA @ NFLX Building a Culture of Analytics Everywhere Tableau Customer Conference 2014.09.09 Blake Irvine Manager, Device Analytics Data Science & Engineering [email protected] Albert Wong Manager, Reporting Platforms Cloud & Platform Engineering [email protected]

Upload: blake-irvine

Post on 25-Jun-2015

599 views

Category:

Data & Analytics


0 download

DESCRIPTION

I presented this at a 2014 Tableau Conference session with Albert Wong. Netflix relies on data to make decisions ranging from buying and recommending content, to improving the streaming experience on devices. This presentation shares our Big Data analytics architecture and the tools used to make data accessible throughout our business, focusing on how Tableau fits into our organization and why it aligns well with our culture.

TRANSCRIPT

Page 1: DATA @ NFLX (Tableau Conference 2014 Presentation)

DATA @ NFLXBuilding a Culture of Analytics EverywhereTableau Customer Conference2014.09.09

Blake IrvineManager, Device AnalyticsData Science & [email protected]

Albert WongManager, Reporting PlatformsCloud & Platform [email protected]

Page 2: DATA @ NFLX (Tableau Conference 2014 Presentation)
Page 3: DATA @ NFLX (Tableau Conference 2014 Presentation)
Page 4: DATA @ NFLX (Tableau Conference 2014 Presentation)
Page 5: DATA @ NFLX (Tableau Conference 2014 Presentation)

Netflix and data in the news...“Giving Viewers What They Want” --New York Times

“The Science Behind the Netflix Algorithms That Decide What You’ll Watch Next” --Wired

Data-Mining Boosts Netflix's Subscriber Base, Showbiz Clout

--AdAge

Page 6: DATA @ NFLX (Tableau Conference 2014 Presentation)
Page 7: DATA @ NFLX (Tableau Conference 2014 Presentation)

BIGDATA

Page 8: DATA @ NFLX (Tableau Conference 2014 Presentation)

Big Data at Netflix

Size● 50+ million members● 1000’s of devices● 100’s of systems● >300B data pipeline events daily● >10B row tables daily

Ubiquitous● Data is everywhere● Many complex systems● Many engineering teams

producing and consuming● Non-streaming teams produce

and consume data● Culturally data driven

Page 9: DATA @ NFLX (Tableau Conference 2014 Presentation)

How do we innovate with Big Data?

Page 10: DATA @ NFLX (Tableau Conference 2014 Presentation)

TOOLSCULTURE

Page 11: DATA @ NFLX (Tableau Conference 2014 Presentation)

Tools

Sting

DATA STORAGE DATA PROCESSORS DATABASE REPORTING

Page 12: DATA @ NFLX (Tableau Conference 2014 Presentation)

Team StructureData Science and Engineering

Marketing

analyst

reporting

data engineering

Finance Product Engineering ...

Business Functions

Page 13: DATA @ NFLX (Tableau Conference 2014 Presentation)

Team StructureData Science and Engineering

Marketing

analyst

reporting

data engineering

Finance Product Engineering ...

NOT WHAT WE DO

Page 14: DATA @ NFLX (Tableau Conference 2014 Presentation)

Netflix Team StructureData Science and Engineering

Marketing

analyst

reporting

data engineering

Finance Product Engineering ...

analyst

reporting

data engineering

analyst

reporting

data engineering

analyst

reporting

data engineering

Business Functions

Page 15: DATA @ NFLX (Tableau Conference 2014 Presentation)

Highly Aligned, Loosely Coupled

Finance

analyst

reporting

data engineering

Marketing

analyst/reporting

data architect

data engineering

Page 16: DATA @ NFLX (Tableau Conference 2014 Presentation)

Highly Aligned, Loosely Coupled

Finance

analyst

reporting

data engineering

Marketing

analyst/reporting

data architect

data engineering

Page 17: DATA @ NFLX (Tableau Conference 2014 Presentation)

Freedom & Responsibility

FreedomDon’t limit accessDon’t limit choicesReduce constraints

ResponsibilityTrustDon’t allow chaosReduce accidents

Page 18: DATA @ NFLX (Tableau Conference 2014 Presentation)

Protected access

CENTRAL DATA

Page 19: DATA @ NFLX (Tableau Conference 2014 Presentation)

Protected access

CENTRAL DATA

NOT WHAT WE DO

Page 20: DATA @ NFLX (Tableau Conference 2014 Presentation)

Unlocked access

CENTRAL DATA

Page 21: DATA @ NFLX (Tableau Conference 2014 Presentation)

OPERATIONAL DATA

Don’t limit choice

CENTRAL DATA LOCAL DATA

Page 22: DATA @ NFLX (Tableau Conference 2014 Presentation)

OPERATIONAL DATA

Don’t limit choice

CENTRAL DATA LOCAL DATA

Sting

Page 23: DATA @ NFLX (Tableau Conference 2014 Presentation)

Why is Tableau a good choice?

QuickIntuitiveRich Visual AnalysisStorytellingEmailed ReportingReusability

Page 24: DATA @ NFLX (Tableau Conference 2014 Presentation)

Who uses Tableau?

Page 25: DATA @ NFLX (Tableau Conference 2014 Presentation)

How do we use Tableau?

Page 26: DATA @ NFLX (Tableau Conference 2014 Presentation)

Examples

● Application build testing● Certification tracking● Operational Excellence

Page 27: DATA @ NFLX (Tableau Conference 2014 Presentation)

Application Build Testing (1/4)

● Team: Product Engineering● Context

○ Application automatically tested at every code checkin

○ Several dozen performance tests run to measure change and avoid regression

● Problem○ Limited graphing tool built into test tool○ Difficult / no customization

Page 28: DATA @ NFLX (Tableau Conference 2014 Presentation)

Application Build Testing (2/4){ "metadata" : {

"TestCaseName" : "Trunk.Rendering.Effects_Mask2","MarkerSetId" : 2472165,"ESN" : "DCQA01","UIBuild" : null,"Build" : "2689","JenkinsJob" : "http://builds.netflix.com/job/208/","Label" : "#2689 / 208","BuildTimestamp" : null,"Changelist" : "2177893"

}, "results" :

[{"MeanFps" : 60.284862537264004},{"MeanFps" : 60.264900662251655},{"MeanFps" : 60.234541577825162}]

}

Build / Test

Page 29: DATA @ NFLX (Tableau Conference 2014 Presentation)

Application Build Testing (3/4)

Page 30: DATA @ NFLX (Tableau Conference 2014 Presentation)

Application Build Testing (4/4)

Page 31: DATA @ NFLX (Tableau Conference 2014 Presentation)

Certification Tracking (1/3)

Team: Certification OperationsContext● We certify the Netflix implementation on many new

consumer electronics devices

Problem● Time consuming to generate insights across multiple

disconnected systems

Page 32: DATA @ NFLX (Tableau Conference 2014 Presentation)

Certification Tracking (2/3)

NTSCertification

Process

Page 33: DATA @ NFLX (Tableau Conference 2014 Presentation)

Certification Tracking (3/3)

Page 34: DATA @ NFLX (Tableau Conference 2014 Presentation)

Operational Excellence (1/3)

Team: Data Science & EngineeringContext● Ensure continuous development does not negatively

impact availability and resilience

Problem● Multiple programs and data sources● Need to link source data patterns to engineering tools

Page 35: DATA @ NFLX (Tableau Conference 2014 Presentation)

Operational Excellence (2/3)

Page 36: DATA @ NFLX (Tableau Conference 2014 Presentation)

Operational Excellence (3/3)

Page 37: DATA @ NFLX (Tableau Conference 2014 Presentation)

Where are we with Tableau?

Page 38: DATA @ NFLX (Tableau Conference 2014 Presentation)

● Netflix is known for being data driven● Big data is available everywhere● Our culture enables analysis everywhere

● Tableau complements our culture● We have organic growth throughout Netflix● Growing part of our reporting platform

DATA @ NFLX

Page 39: DATA @ NFLX (Tableau Conference 2014 Presentation)

Blake IrvineManager, Device AnalyticsData Science & [email protected]

Albert WongManager, Reporting PlatformsCloud & Platform [email protected]

What can we answer?