The Big Data Canvas

Swire Leadership Forum

Mart van de Ven | m@droste.hk

most definitions of

HUGE, OIL, HADOOP, MONEY, BUSINESS TRANSFORMATION

Big Data

Big Data

an honest definition of

a stand-in for all the ideas which apply machine intelligence to datasets to generate business value

 

Big Data

Big Data

The DIMO Model​

Magic Out

Data In

Thinking about Big Data

Focus on Requirements

Collection

Storage

Application

Adoption

Thinking about Big Data

Focus on Applications

Knowledge Discovery

Decision Support

Data Driven Services

Reporting

Thinking about Big Data

Focus on Value

PROJECT MEGA

Mall Experience Great Again

Big Data Canvas

Objectives

Big Data Canvas

Objectives

Sell 200,000 units

Market Leader in Malaysia

iPad Connectivity in sales booths

Fire Jimmy and then..

Measurable

Unbounded

Inspirational

Many objectives

Data Plan ...

... not Data Problem

Actionable

Includes the solution

Reduce tenant turn-over by providing value-added business intelligence services

MALL EXPERIENCE GREAT AGAIN

DROSTE

NOV '17

1.0

Big Data Canvas

Success Criteria

Big Data Canvas

Success Criteria

Why don't we get this right?

Data Availability

Attribution

Metrics are Proxies

Reduce tenant turn-over by providing value-added business intelligence services

MALL EXPERIENCE GREAT AGAIN

DROSTE

NOV '17

1.0

No turnover in top 10 tenants, lowest overall turnover rate in the market

Big Data Canvas

Benefits

Big Data Canvas

Benefits

Defining benefits in the light of uncertainty:

Develop scenarios

Beyond monetary 

Iterate with the Canvas

Reduce tenant turn-over by providing value-added business intelligence services

MALL EXPERIENCE GREAT AGAIN

DROSTE

NOV '17

1.0

No turnover in top 10 tenants, lowest overall turnover rate in the market

(1) Avoid 40% of est. US$12M losses due to tenant turn-over

(2) Value-add USP 

(3) Opt. mall utilization

Big Data Canvas

Data

Big Data Canvas

Data

Source

Temporal Dimension

Feature Engineering

Accessibility

Internal

External

Static

Dynamic

Structured

Unstructured

Access Granted

Siloed or Propietary

Reduce tenant turn-over by providing value-added business intelligence services

MALL EXPERIENCE GREAT AGAIN

DROSTE

NOV '17

1.0

No turnover in top 10 tenants, lowest overall turnover rate in the market

(1) Avoid 40% of est. US$12M losses due to tenant turn-over

(2) Value-add USP 

(3) Opt. mall utilization

Mobile signals

CCTV footage

Sales type & volume

Employment & Tourism data

Perf. other malls

Big Data Canvas

Constraints

Big Data Canvas

Constraints

Regulatory Compliance 

Resources

Business Context

Data

Privacy

Discrimination

Hardware

Software

Speed

Quality of Service

Accessibility

Quality

Reduce tenant turn-over by providing value-added business intelligence services

MALL EXPERIENCE GREAT AGAIN

DROSTE

NOV '17

1.0

No turnover in top 10 tenants, lowest overall turnover rate in the market

(1) Avoid 40% of est. US$12M losses due to tenant turn-over

(2) Value-add USP 

(3) Opt. mall utilization

Mobile signals

CCTV footage

Sales type & volume

Employment & Tourism data

Perf. other malls

Privacy / Identification

Sensor Density

Access to POS

 

Big Data Canvas

Risks

Big Data Canvas

Risks

Managing Risk :

Identify potential sources

Evaluate consequences

mitigation strategies

Reduce tenant turn-over by providing value-added business intelligence services

MALL EXPERIENCE GREAT AGAIN

DROSTE

NOV '17

1.0

No turnover in top 10 tenants, lowest overall turnover rate in the market

(1) Avoid 40% of est. US$12M losses due to tenant turn-over

(2) Value-add USP 

(3) Opt. mall utilization

Mobile signals

CCTV footage

Sales type & volume

Employment & Tourism data

Perf. other malls

Privacy / Identification

Sensor Density

Access to POS

 

Tenants not responsive to intelligence

BI cannot drive enough value to offset rent pressure

Big Data Canvas

3 Take Aways

3. Big Data isn't special.

1. Big Data isn't Magic. 

2. Success is 70% strategy.

... you might be.

... eyes on the prize.

... and 30% luck.

Thank you.

Please keep your questions

for the Q & A 

Principal Data Scientist , Droste

Mart van de Ven

m@droste.hk