Agile BI success factors

Success Factors

Jean-Michel Franco Innovation & Solutions Director

[email protected]

Telephone, : +33 6 67 70 01 32

Twitter : @jmichel_franco

Agile Business Intelligence

Business & Decision is a global

Consulting & Systems Integrator

2012 : 221,9 M€

2

2 500 Employees 16 Countries Multi-Specialist

BI

PM

CRM EIM

E-bus

Expertise recognized by thought leaders, Software vendors and industry analysts

• Business Intelligence & EPM “European Marketscope for BI Services”. Gartner

• Customer Relationship Mgt & MDM “CRM Wordwide Magic Quadrant”. Gartner

• E-Business “Interactive Design Agency Overview, Europe, 2013 ”. Forrester

3

BI: raising expectations from Lines of Business …

Source : Gartner Survey Analysis: CFOs' Top

Imperatives From the 2013 Gartner FEI CFO

Technology Study

4

…while IT ’s ability to deliver on promises is being challenged

Source : Gartner Survey Analysis: CFOs' Top Imperatives From the 2013 Gartner FEI CFO Technology Study

Innovating through IT, close to the field

• Discover : raising awareness on emerging

technologies and use cases

• Incubate : a proof of concept based

approach to experiment IT in context of

each business process

• Productize once proof of concept has

been made

• Continuously improve : extend existing

environment rather than replace them -> a

lean approach to innovation, by increments

…

• Shares lessons learned, turn « next

practices » into « best practices ».

•

5

http://blogs.hbr.org/cs/2012/03/look_to_it_for_process_innovat.html

6

Top down

approach:

Enterprise

BI

Bottom up

approach :

Personal

BI

Management teams

Is Business Intelligence in midstream ?

Enterprise BI as we know it

Occasional user 70+ %

“advanced” user: 30- %

7

Enterprise BI as we want it

8

9

BI as we want it: Success factors

People

Organi-zation

Metho-dologies

Tools

Infra-structure

Business/ processes Analytics

Data governance Information Management

Data Discovery Self Service BI

Self Service Information Management

Data Lab : environment for prototyping and self service

access to data

Close to the field : a front office to collect ideas,

experiment and design + back office to roll out on

a wide scale

Upstream collection of business needs

Template based agile methodologies

The technology layer

10

11

The people dimension

Socialize Business Intelligence or Changer gravity of Business Intelligence

To engage Lines of business beyond the project blueprint phase

(Model design, shared system of measurement, business glossaries…)

12

Infrastructure dimension : the Data Lab principle

Enterprise BI

Data Warehouse

Data Mart

Packaged apps,

Dash-boards Self Service

Data Lab

Ephemeral stores

Application prototypes

Self-Service

Sanctioned data

Shared analytics

Enterprise level models

Sanctioned Data sources

Unsanctio-ned data

13

Project dimension: Rethinking the entire BI lifecycle

When Challenge Solution

Before the BI project Identify emerging business needs. Formalize business cases. Prove the concept.

Bring the tools close to the use cases at early steps. Incubate new technologies. Identify key user and empower them

During the BI project White page syndrome Difficult to validate design, to anticipate problems (ex : data quality).

Agile methodologies Template based design

After the BI Project roll-out

Evolve the system « on the fly » Establish a self service usage

Empower a certain category of business users to: - accompany and coach - Manage data governance - Identify change of business needs

Business Objectives

Company is best in class in terms of water

quality and aspires to strengthen this

leadership

Project 'Water Quality Performance' aims to

provide the platform to drive future

performance in that area

Chosen approach • IT empowers business users

(Statisticians) to get knowledge

out of external data and allow

cross analysis with internal data

• Agile approach :

Establishing agile BI before projects ; example in utilities

– Ability to source

external “multi-

structured “ data

14 million rows at

that time

– Allow data

crunching

(including quality

checks) and

analytics

– Timing : 1 month

before first results

– Proof the concept

on a small scale

before wider roll-

out

– “show the data”

first, then learn

and refine the

design to adjust

the solution to the

business need

Business objective

Re-engineer the marketing system

foundations :

Chosen approach

• Leverage a standardized data model

(Acord) covering the 17 business

domains of insurance

• iterative and incremental design

approach on three areas:

Agile during the BI project: Example in insurance

– Customer master data and

marketing data warehouse

– Customer analytical Data

Mart (scoring,

segmentations…)

– Packaged software for

multi-channel marketing

campaigns (Neolane)

– Data Modeling (2 weeks sprints for

each considered data

domains)

– Data integration

– Data quality

assessments and

audits

15

Agile BI all along the BI initiative : eexample in Life Sciences

Business objectives

Relaunch Business Intelligence

initiatives :

Chosen approach

– Solidify the information back

office (data models, shared

master data, data quality &

governance)

– Closely match Business

Intelligence to the need of

each line of business

– Better catch business needs

upstream and downstream

(before and after project

launch)

– Take advantage of data

discovery and data

visualization tools

Catch

Business

needs

Design

Productize

Key user, at each lines of business, to collect business needs and autonomously discover the data

Prototyping at very early steps of each project

A center of expertise and shared standards to quickly roll out and globalize BI initiatives

Drive

usage

Well defined organizations to accompany BI usages and make sure of the efficient usage of data

16

Success Factors

Jean-Michel Franco Innovation & Solutions Director

[email protected]

Telephone, : +33 6 67 70 01 32

Twitter : @jmichel_franco

Agile Business Intelligence