Data Science for Growth

Slide 1

Slide 1 text

Data Science for Growth Gaza Sky Geeks Sean Patrick Braithwaite

Slide 2

Slide 2 text

• Data Science as a Discipline • Data Science as a Process • Techniques for Growth • Understanding Users • Understanding Groups • Understanding Change Agenda

Slide 3

Slide 3 text

Data Science as a Discipline

Slide 4

Slide 4 text

Data Science as a Mix of Disciplines

Slide 5

Slide 5 text

Data Science as Hype

Slide 6

Slide 6 text

Data Science is a Process

Slide 7

Slide 7 text

Data Science is an Active Process that drives Incremental Change

Slide 8

Slide 8 text

Navigation goes Wrong

Slide 9

Slide 9 text

Data Science is a Process that helps navigate building products

Slide 10

Slide 10 text

Data Science is a Process that makes other processes Efficient

Slide 11

Slide 11 text

When to use Data Science

Slide 12

Slide 12 text

Data Science for Growth Speed & Efﬁciency, Speed != Efﬁciency

Slide 13

Slide 13 text

Small data has Challenges

Slide 14

Slide 14 text

Small data has Advantages

Slide 15

Slide 15 text

Techniques for Growth

Slide 16

Slide 16 text

Understanding your Users Username Age Channel $ Spent Active Omar 21 Twitter 2 Yes Nour 45 Facebook 4 No Fatima 30 Organic 7 Yes … … … … …

Slide 17

Slide 17 text

Understanding your Users Username Age Channel $ Spent Active Omar 21 Twitter 2 Yes Nour 45 Facebook 4 No Fatima 30 Organic 7 Yes … … … … … as mixes of Numeric and Categorical ﬁelds Numeric Categorical

Slide 18

Slide 18 text

Understanding your Users Username Age Channel $ Spent Active Omar 21 Twitter 2.4 Yes Nour 45 Facebook 4.2 No Fatima 30 Organic 2.7 Yes … … … … … as mixes of Observed and Derived ﬁelds Derived Observed

Slide 19

Slide 19 text

Understanding your Users through plotting as Vectors Age $ Spent Channel Fatima Twitter Organic Facebook

Slide 20

Slide 20 text

Understanding your Users through Correlation Age $ Spent Channel Twitter Organic Facebook

Slide 21

Slide 21 text

Understanding your Users Through Exploring Correlations $ Spent orders Cost Impressions Clicks $ Spent orders Cost Clicks Impressions

Slide 22

Slide 22 text

Understanding your Users as Groups of Users Campaign $ Spent 1 2 3 4 1 2 3 4 5 6 Acquisition Cost $

Slide 23

Slide 23 text

Understanding your Users In practice, using tools • Excel: Surprisingly powerful! Works at small scale • Mixpanel: Cheap to get started, straight forward integration • IPython: Learning curve, works well at small scale (http://python-for-multivariate-analysis.readthedocs.io/)

Slide 24

Slide 24 text

Understanding Diﬀerences with Multivariate Testing Or

Slide 25

Slide 25 text

Understanding Diﬀerences with Multivariate Testing B A\

Slide 26

Slide 26 text

Understanding Diﬀerences in the outcomes of experiments Null: Option A and option B are equally good Alternative: Option B is signiﬁcantly better than option A

Slide 27

Slide 27 text

Understanding Diﬀerences that are Significant Conversion Rate Probability density 0.15 0.225 0.1 Probability > 0.19 0.19 Most likely conversion rate 5%

Slide 28

Slide 28 text

Understanding Diﬀerences In the Distribution of performance

Slide 29

Slide 29 text

Understanding Diﬀerences By understanding p-values: Conversion Rate P(Conversion Rate) Observed Conversion: 0.19 Expected Conversion: 0.167

Slide 30

Slide 30 text

Understanding Diﬀerences Avoiding problems • Respect assumptions: Independent, normally distributed • Collect enough data: Low powered tests • Test as necessary: A-A tests when cost of being wrong is high

Slide 31

Slide 31 text

Understanding Diﬀerences In practice, using tools • Google Analytics: Basic free • Optimizely: More features, paid • Custom: Necessary at scale, expensive to build/maintain (http://www.evanmiller.org/ab-testing/)