Sander Klous - Op de 1e editie van Data Donderdag

Data analytics

Sander Klous

Big Data Analytics Lead

[email protected]

1 © 2013 KPMG Advisory N.V., registered with the trade register in The Netherlands under number 33263682,is a subsidiary of KPMG Europe LLP and a member firm of the KPMG network of independent member firms affiliated with KPMG International Cooperative (‘KPMG International’), a Swiss entity. All rights reserved. Printed in The Netherlands. The KPMG name, logo and ‘cutting through complexity’ are registered trademarks of KPMG International.

Contents

Process

■ Approach

■ Big Data awareness event

■ Agile development

■ Evaluation

Examples

■ Life changing events

■ Financial health

■ Hospital care path optimization

■ Data driven decision making

■ Dynamic influencer modeling

■ Crowd monitoring and event security

■ In store customer buying patterns

Process


Our approach for a value added result The process of achieving a proof of concept for the selected cases is aligned with the possibility to develop business cases for the other opportunities

Insight into digitalizationDiscovering the digital possibilities

Breaking the rules

Related worlds

Ambition

Ambition

Customer perspective

Powertrackplanning

Business value

Ideas for digitalizationVision Generating an appetite

Preparation for and creating client cases using a Big Data opportunity event Proof of Concept for selected cases and

hand over Transformation of the solution

and business model

Envisioning…

uncovering client cases using a Big Data opportunities

Preparing…

selected client cases for Proof of Concepts

Reviewing…

evaluating the results and determining the

way forward

Developing…

the Proof of Concepts of the

selected client cases

Coordinating the project using a phased and agile approach

Continuous interaction to orchestrate Big Data strategy process and if necessary

adjust business case directions.

2 weeks 10 weeks to be decided

Part of this proposal

Additional activities

Client cases


Preparation for and creating client cases using a Big Data awareness event Stakeholder commitment and involvement are key in the activities that proceed to the Big data awareness and opportunity event

To achieve the mutually defined results during the Big data awareness session, preparation is key. This phase consists of multiple activities to prepare the necessary input for the awareness session. In addition to these activities, client stakeholder commitment and involvement are key success factors during this phase.

7: Opportunity event preparation

Conducting an extensive analysis that aims to identify relevant trends in the financial services industry and big data initiatives.

2: Forward Thinking

Identifying relevant business models in related industries will provide valuable learning experience as well as identifying future opportunities.

3: Related Worlds

A customer analysis identifies future needs and priorities that contribute in relevant customer profiles.

4: Customer Perspective

Based on the strategy, trends, and customer profiles analyses, the main value drivers are determined and will form the basis for potential business or revenue models.

5: Business Value

To realize a shared vision a strategy analysis is required and includes an understanding of corporate and e-strategy, a competitive benchmark, and portfolio analysis.

1: Shared vision

Creating client cases by stakeholders using an outside-in view on the possibilities of Big Data for the client business model and strategy.

8: Big Data opportunity event

Adding facts to the fact wall during the preparation will ensure that all insights are used during the event that contributes to client’s future business model.

6: Fact wall


Pilot for selected cases and hand over Based on the selected client cases, working pilots are delivered in an agile approach

Scenario I. Using the Teradata EDW Scenario II. Using a Hadoop cluster Scenario III. Using a Teradata Aster cluster

Based on the selected client cases, the required Big Data environment is deployed. There are, for example, three scenario’s: 1. Using the existing Teradata EDW 2. Using a Hadoop cluster 3. Using a Teradata Aster cluster

1: Deploying the Big Data environment

The Scrum Team has to get ready and the Product Backlog has to be groomed. Refined client cases serve as the starting point for development.

2: Getting ‘Ready’

The Scrum Team and Product Owner develop working prototypes in 4 Sprints. Each Sprint starts with a Sprint Planning Meeting and end with a Retrospective. Our agile approach allows for changing requirements throughout the development cycle and ensures getting maximum value from the Big Data environment

3: Delivery using Scrum


Pilot Retrospective

When the prototypes have been used for a couple of weeks/months, a pilot Retrospective will be held:

1. The value of the pilot will be evaluated. What did it contribute? What have we learned? This concerns both technical as well as organizational findings.

2. What is the future path for the client concerning Big Data? What are the next steps? How can we leverage data as a strategic asset? What is needed to make us a data-driven company?

The pilot Retrospective is used as input for development of the Big Data Strategy and allows for uncovering Big Data opportunities that maximize business value in support of the client's proposition or uncover possibilities for business model innovation.

Transformation of the solution and business model Evaluation of the pilot and defining the future path...

Using our proven assessment methodology, we will evaluate the pilot findings together.

1: Pilot review

The possibilities of Big Data for the client will be re-assessed. How do we leverage the value of our data? What actions are needed?

2: Future path Transformation of the solution and

business model


Weeks (dates to be discussed)

1/2 3/4 5/6 7/8 9/10 11/12 13

8-12 Key players (Client)

1 SME (KPMG)

2 Facilitators (KPMG)

2 Data scientists (KPMG)

1 Project Manager (KPMG)

1 Teradata expert

1 Product Owner (Client)


1 Visualisation expert (KPMG)

1 Project Manager/Scrum Master (KPMG)

1 Teradata expert

■ End product

■ Documentation

■ Roadmap

Envision Prepare Develop

Client Big Data strategy development

■ Kick-off meeting

■ Awareness session

■ Workshops

■ Sprint Planning meetings, Demo’s, Retrospectives

Review

■ Workshop

■ Product Backlog ■ Retrospective

1 Product Owner (Client)


1 Visualisation expert (KPMG)


1 Teradata expert

■ Selected client cases

■ Organizational awareness

■ Working product increments (two week cycle)

■ Basic Big Data environment

■ Refined client cases

■ Ready Product Backlog and Ready Team

Activ

ities

R

esul

ts

Peop

le

4 Key players (Client)

1 SME (KPMG)

1 Facilitator (KPMG)

1 Data scientists(KPMG)


1 Teradata expert

Project management

Go / No Go

Go / No Go

Go / No Go

Timeline Approach is characterized by clear steps and focuses on delivery in 12 weeks


Our DNA Service as a Data Institute in an Eco-System

• Output via API/ custom visualization

• Entire (custom) solutions for (multiple) clients

II. Crowd & Mobility Management

Storm Cluster Real-time processing

2a

1

Input Data

Monitoring: G

anglia

Configuration m

anagement: Puppet

Cloudera Hadoop Cluster

Batch processing

2b

MongoDB Cluster 3

4

Rest Interface

5a 5b

Widgets

Varnish Cache Varnish Cache 6

Load Balancer

User Request

1’ Web Apps

Service Layer

Security

KPMG Big Data Analytics Platform Optimize time to market

Backend

• Technologies employed

• Based on Open Source resources

• Flexible and scalable solution

Examples

For internal use only


Personalized Offerings and Pricing Life-changing events

Combine the internal data of a bank with publicly available data from Funda.nl

Data Hub Customer Analytics Institute

Company A

Company B 1

3

3

2

encr

yptio

n

encr

yptio

n

&


Risk Management Financial Health

Pattern recognition in income and expenses – Financial Health predictions

Parameters a & b towards Financial Health Dept.

Parameters a & b towards Non-Financial Health Dept.

Expenses within the range for sustainable behavior

Customers who were never considered by the Financial Health Dept. balanced their spending with their income.


Monte Carlo simulation to analyze operational efficiency of Operation Room planning at a hospital.

Operational Excellence Hospital care path optimization

Sim

ulat

ion

Rea

lity

2012 2014


Customer Profiling and Operational Efficiency Decision-making based on generated revenue

The most popular handset types used in the entire Netherlands versus Amsterdam. The information is based on the revenue per antenna from analysis of the call detail records. Note that Samsung is most popular in general, but in Amsterdam, the trendier Apple’s iPhone is mostly used instead.


Customer Profiling Dynamic Influencer Modeling in Social Media Networks

Dynamic Influencer Model of Business on Twitter

Dynamic Influencer Modeling is used to mine social networks for their influential members. It calculates the influencers according to both the formal and informal connections. It computes the influencers dynamically, computing the influencers in real-time using the output of Tribes identification information. Each of the connections is weighted dynamically, to create ranking of members for diverse topics, such as “Toyota Prius”, “fuel economy” and “baseball draft”.


Location Aware services Crowd monitoring and event security during the coronation

KPMG Managed Service Big Data Analytics platform

Analytics as a Service

(AaaS)


Location Aware services In store monitoring: understanding customer buying patterns

Background

Tracking customer behaviour is important information which can be gathered using several types of sources. WiFi signals and old-school security cameras are for instance used more and more to store shopping patterns from customers.

Benefits

■ The goal is to define which variables affect a purchase and make changes accordingly. For example, a store might deploy more salespeople, alter displays or put out red blouses instead of blue, according to a story on Bloomberg

Challenges

■ Storing WiFi signals for analysis.

■ Analyzing the customer behavior and plotting patterns, graphs and trend lines.

■ Storing sales data, or motion sensors for extra information about certain products.

■ Storing other relevant (social Media) information.

Solution

■ The analysis of any of these separate data streams will provide vital information about the shopping behaviour of customers. Combining all the data creates even richer information.


Location Aware services Live dashboards

Service Catalog Capture rate ■ How many people that walk by actually enter

Dwell time ■ How long does an average visitor stay inside

Conversion ■ How many customers that enter buy a product

Visits ■ How many people visit and how does this vary over stores

Returning visits ■ How often do customers visit a store

Heatmaps ■ What is the most visited area in a store

Benchmarking ■ A/B testing between different locations and dates

Store Layout Optimization ■ What is the best store layout

Trends ■ Above indicators over time

Customer Benefits Location based loyalty ■ Reward Programs (bonus miles) ■ Customer Engagement levels (bronze, silver, gold)

Location aware discounts ■ Based on customer loyalty

In store navigation ■ Help customers find their preferred store path

Queue time optimization ■ Minimize waiting time at cash registers or counters

Hottest item identification ■ Help customers identify best selling products

Collaborative filtering ■ Suggest new products related to customer interest

Personalized offerings ■ Personalized offerings based on better customer profiles

Customer journey ■ Help customers with their purchase decision in stores and at home


Location Aware services Heatmap of CHEP 2013 (Conference in Beurs van Berlage)

© 2013 KPMG Advisory N.V., registered with the trade register in the Netherlands under number 33263682, is a subsidiary of KPMG Europe LLP and a member firm of the KPMG network of independent member firms affiliated with KPMG International Cooperative (‘KPMG International’), a Swiss entity. All rights reserved. Printed in the Netherlands.

The KPMG name, logo and ‘cutting through complexity’ are registered trademarks or trademarks of KPMG International.