Building Data Culture

Building a Data CultureIsaac Souweine

Head of Product @ Frank & Oak Presented to Startup Next, Montreal

March 23, 2015

Personal Intro● Columbia (Religion)● Scholastic● INSEAD● Yahoo● Pollenizer● Frank & Oak

This Talk

DATA AVAILABILITY

DATA CULTURE

AWESOME CO.

Part 1 - Data Availability

The F&O BI Stack (Simplified View)CUSTOM ANALYSIS

VISUALIZATION + ANALYSIS

DATA WAREHOUSE

EVENT PROCESSING

EVENT TAGGING \

AUDIENCE ANALYTICSDATA

Making Data Availability Work

● Right tech● Right people● Right attitude● Right execution

IF A USER CLICKS A BUTTON BUT AN EVENT ISN’T TRIGGERED, THEN DID THE SERVER CALL GET MADE?

Part 2 - Data Culture

The 3 Pillars of Data Culture

1. A shared belief that a product experience can be distilled into data points

2. Widespread agreement on what those data points are and how they are computed

3. A company culture that puts those key data points at the center of ALL decision making

Making Data Culture Work

● Data availability!!● Less is more - OMTM● Unit economics● Cohorts are king● Empower the end user● Testing methodology● Unified systems● Data as a necessary condition

Data Culture in Practice

● Pull Dashboards● Push Dashboards (email)● Metrics reviews● Vigorous debate● Visual data culture (screens)

Example - Data Questions

● Which traffic sources deliver the best CAC?● Which emails deliver the best CTR?● Which version of the HC Join page converts best?● How effective are modals in driving feature

adoption?● Is the increased AOV for Select Crates cancelled

out by the higher return rate?

A final thought. . .

[email protected] @sonofsarah