customer insight and prediction for b2b marketing 22nd ... · •revenue in the test groups is up...

A SELECT INTERNATIONAL COMPANYwww.sv-europe.com

Customer Insight and Prediction for B2B Marketing22nd November 2013

A SELECT INTERNATIONAL COMPANY

Agenda• Arrival, registration and coffee

• Introduction: what is predictive analytics and is it realistic for you?

• Typical predictive analytics applications: where is it being used, for what and with what results?

• Break

• Analytics in action: Roger Watson – Chubb Fire and Security

• Technology overview: demonstration of tools and techniques

• Getting started: what to consider and how to begin

• Close and Lunch


Aims and Objectives

• Introduce Smart Vision, IBM and predictive analytics

• Share practical approaches to using advanced analytics for improved business to business marketing and customer management

• Encourage discussion and debate with us and within the group

• De-bunk some myths and set expectations

• Suggest practical approaches for getting started


About Smart Vision…

• Premier IBM Partner• Accredited Support Providing Partner for SPSS• Specialist in SPSS analytical products and BI

– Currently resourcing the IBM SPSS public training schedule

• Staffed by former SPSS employees– Former UK General Manager– Former SPSS Brand Leader, EMEA– Former UK Head of Professional Services– Senior Professional Services Specialists

• Experience in advanced analytics– B2B, retail, telecommunications, leisure and hospitality, financial services,

publishing, utilities…

A SELECT INTERNATIONAL COMPANYhttp://www-01.ibm.com/software/data/2012-conference/awards.html

http://www-01.ibm.com/software/data/2012-conference/awards.html


Projects in the last 12 months….


Predictive analytics helps connect data to effective action by drawing

reliable conclusions about current conditions and future events.

— Gareth Herschel, Research Director, Gartner Group

What is predictive analytics?


Core capabilities driven by advanced predictive analytics

grow risk

attract

retain

BehaviouralDescriptive

InteractionSentiment

When?

How frequent?

No. contacts?

Feedback?

SIC, Credit Rating,

Employees, T/O,

EBIT...

Offers?

Payment history?

How often?

How long?

What value?

Feedback/Surveys

Social Media


Types of predictive modelling

Propensity/

ClassificationClustering

Association / Sequence

Time Series

Identify groups

within a

population

displaying

homogeneity

(based on a wide

array of data)

Identify

repeatable

patterns of

behaviour or

sequence

Forecast a

future value

over a defined

time period

Predict a

particular type

of outcome


Types of predictive modelling• Classification / Propensity

– What customer account is most likely to respond / convert / cancel based on historical response data and the array of behavioural data we have about them?

• Clustering

– How can I divide my account base into meaningful and usable groups as a framework for marketing communications? By value, by service or product portfolio

• Association and Sequence

– What is the optimal sequence and frequency of events and interventions that lead to a one off purchase from a company becoming a high value account?

• Time Series

– What will product demand be next month / quarter / year?

• Optimise product sourcing decisions

• Identify pricing errors during product lifecycle

• Predict end of life products in time to divest stock held


Is it practical for your business?

• Do you need to be statistically trained?

– No…

– Appropriate skills can be transferred and learnt

– Business understanding is just as important as technical skill

– Analytical skills can be taught

• Data readiness?

– You need some data, certainly…

– Does not need to be organised into a tidy single supporter view

– Data is often [always] fragmented and disparate (occupational hazard)


Is it practical for your business?• How do we get started?

– You can start small and develop in phases

– Demonstrate initial ROI before releasing further investment

– Plan your first projects using a recognised analytical methodology

• Cost to get started?

– Approximate, minimum investment £20 – 30k

• Key analytical tools implemented

– Data access, management and manipulation

– exploratory analysis, profiling and modelling capabilities

• Skills transfer and knowledge to continue independently

• First project (s) completed

• Actionable results, deployed in the business


John McConnell - Services

Typical

predictive analytics


Interest in predictive analytics

‘Predictive Analytics’

‘Business Intelligence’


So who uses predictive analytics?

• Commercial Sector


• Not-for-profit sector

So who uses predictive analytics?


Predictive customer analytics: driving actionable insight

Acquire customers: • Understand who your best customers are

• Connect with them in the right ways

• Take the best action maximize what you sell to them

Grow customers: • Understand the best mix of things needed by your customers and channels

• Maximize the revenue received from your customers and channels

• Take the best action every time to interact

Retain customers: • Understand what makes your customers leave and what makes them stay

• Keep your best customers happy

• Take action to prevent them from leaving


Predictive operational analytics: driving operational efficiency

Manage operations: • Maximize the usage of your assets

• Identify the impact of investment

• Ensure inventory and resources are in the right place at the right time

Maintain infrastructure: • Understand what causes failure in your assets

• Maximize uptime of assets

• Reduce costs of upkeep

Maximize capital efficiency: • Improve the efficiency and effectiveness of your assets

• Reduce operational costs

• Drive operational excellence in all phases: procurement, availability and distribution


Predictive threat and fraud analytics: driving mitigation and prevention

Monitor environments: • Identify leaks

• Increase compliance

• Leverage insights in critical business functions

Detect suspicious activity: • Identify fraudulent patterns

• Reduce false positives

• Identity collusive and fraudulent merchants and employees

• Identify unanticipated transaction patterns

Control outcomes: • Take action in real-time to prevent abuse

• Reduce Claims Handling Time

• Alert clients of transaction fraud


Typical Application Aims

Grow Defend

Market

Share

Profit

Acquire More Customers

Build a Reputable Brand

Anticipate Demand

Maximise Satisfaction

Lower Cost of Acquisition

Cross Sell

Up Sell

Maximise Lifetime Value

Maximise Loyalty

Address Poor Satisfaction

Lower Churn Rates

Reactivate Customers

Minimise Defaults

Prevent Fraud

Prevent Waste

Maintain Availability


Typical predictive analytics applications

• Predictive modelling

– Marketing Response

– Customer Acquisition

– Cross-Sell/Up-Sell

– Customer Retention

– Asset Failure

– Fraud Detection

– Satisfaction Modelling

• Segmentation– Cluster Analysis

– Life Time Value

– Loyalty

– Store Clusters

• Other applications– Basket Analysis

– Forecasting

– Sentiment Analysis

– Anomaly Detection


What sorts of people use predictive analytics?

• Marketing Managers/Analysts

• Fraud Officers

• Insight Directors

• Segmentation Strategists

• Asset Risk Managers

• Data Scientists

• Market Researchers

• Assortment Managers

• Credit Risk Officers

• CRM Analysts

• Information Service Managers

• Heads of MI

• Data Mining Consultants

• Head of Supporter Relationships

The most effective instigators of predictive analytics are

very often data-literate, business-focused people


Brammer

• Leading distributor reduces cost of carrying surplus stock and improves customer service

• Applications and Benefits– IBM/SPSS predictive analytics helped Brammer to manage its inventory more

efficiently, significantly reducing the need to carry surplus stock, resulting in a total inventory reduction of £31.1 million in one year

– Inventory turnover improved from 3.2 times at the end of 2008, to 3.7 times at the end of the first half of 2009

– Greater understanding of patterns and trends in customer purchasing data helps Brammer forecast marginal stock products more accurately and improve customer satisfaction by making a wider product range available for immediate dispatch

– Detailed insight into inventory requirements has helped Brammer develop closer relationships with strategic suppliers leading to further cost benefits


Case Study: Cablecom

• Based in Switzerland

• Core business: cable TV

• Diversified into:

–Broadband internet access

–Digital phone

–Pre-paid mobile

• Business issue: retention of broadband customers

–High cancellation rate at end of initial contract


Identified crucial point in lifecycle

• Created customer satisfaction survey, run at month 7 of initial contract

• Ten “0-10” questions, one free text

• Converted to single “satisfaction score”, 0-100

–100 = perfectly satisfied

– 0 = totally dissatisfied


Predictive approach

• Combined satisfaction scores with other data assets:

–Demographics

–Product ownership and usage behavior

• Built models to predict satisfaction score for all customers

• Used predictive satisfaction score to drive retention marketing

• Result: churn reduced from 19% to 2% in treated group


Cablecom example 2: NPS plus text mining

• Net Promoter Score (NPS)

“On a scale of 0-10 how likely is it that you would recommend Cablecom to a friend or colleague?”


Combine NPS (KPI) with

Customer Feedback (key to actions) and Text Mining

Do more of these things

Supplementary open ended question:- Promoters

“what is the single most important thing that makes you likely to recommend us?’’

- Passives“ what is the single most important thing we could do to make you more likely to recommend us?’’

- Detractors “ what is the single most important thing that currently makes you unlikely to recommend us?”

Fix these things


Results• Macro level:

– Information from supplementary question, categorised, reveals areas for improvement (product owners, market managers, operations managers)

• Micro level:

–Customer issues drive 1:1 interactions and resolutions

• In 3 months:

–Satisfaction improved in > 50% of cases

–23% of detractors converted to promoters


Will you buy a car today?

• Fiat identifies the most likely customers and prospects

• Applications and benefits

– Improved customer response rate to marketing initiatives by 15-20 percent.

– Improved customer loyalty by 7 percent.

– Supports continuous improvement of dealerships and repair facilities.

– Centralized analytical reporting and modeling system enhances productivity and lowers costs.

– Efficiently works with large Oracle database

• containing history on 64 million customers


Predictive maintenance snapshots

Major helicopter manufacturer Individualized maintenance plans for each

helicopter based on history and operations Enables “just in time” inventory management

Manufacturer of mining equipment

Proactively identified problems and the best action before failure

Saved $1 million in repair costs in under 2 weeks

12-14x ROI (return on investment) in just 4 months

UK water companies Reduce internal flooding incidents by predicting

asset failure risk

Predict risk of pollution events and take pre-emptive action


Retaining subscribers

• Annual Magazine Subscription Renewal Modelling• Predict the likelihood of each customer to renew at their next renewal

• Ensure predictive accuracy

• The model must make sense to the business – it must be usable and ‘deployable’

• Pilot ran across three major lifestyle titles


Data sources and fields

De

scri

pti

veCompany/Individual

Company size

Business type

Job function

Age

Association membership

Gender

LocationV

alu

e Lifetime

Lifetime value

Annualised value

Back issue claims

Payment method

Time taken to pay

Amount paid last time

Mar

keti

ng Frequency of contact

Acquisition Channel

Renewal channel

Subscription term

Preferred response method


Project Context and Results

• Challenges around gaps in data– Created proxy measures for annualised value

– Refined range of date used through EDA

• Analysis– Decision trees used because…

• Good accuracy (tested with a “holdout” sample)

• Visual representation

• Deployment– The winning model was deployed as a test/control 50:50 split

Business Understanding

Data Understanding

Data Preparation

Modelling

Evaluation

Deployment


Results

• Revenue in the test groups is up 18%• Profit in the test groups is up 21% • The success of this test means it is being rolled out across

100% of records for participating titles• Approach implemented on further titles

– We are working collaboratively to further automate the model development and deployment process


Challenge:

In a highly competitive environment, Banco Itau Argentina wanted to find a more

effective way to increase customer satisfaction and lifetime value while maximizing the

bank’s profitability.

Solution:

The bank chose IBM Business Analytics software and used it to create predictive

models that identified clients with a high probability of accepting marketing offers. It

also used sales optimization technologies to maximize the effectiveness of its

marketing campaigns.

Impact:

Improve understanding of the customer to provide an unprecedented level of

targeting and coordination of campaigns which decreased cost and increased

revenue

Implement 20-30 rolling sales campaigns

Increase revenue from existing clients by 40 percent

Increase the total retail customer contribution margin by close to 60 percent

Banco Itaú Argentina

Optimizing customer cross-selling and acquisition strategies

“We wanted to increase

customer satisfaction and life-

time value while maximizing

bank profitability through the

analysis and execution of

optimal cross-selling and

acquisition strategies. That’s

why we chose solutions from

IBM.“

Mauricio González Botto,

COO, Banco Itaú Argentina


American Public University SystemUsed predictive analytics to boost student retention and academic excellence

“For the first time, we have a statistical

model that can show, based upon data,

which students are most susceptible to

attrition. Before we started using IBM

SPSS predictive analytics, we were just

guessing from among hundreds of

variables and trying to put them together

by hand.”

- Dr. Phil Ice, Director of Course Design

for APUS.

The need:

APUC wanted to boost student retention and academic excellence by better predicting

the likelihood of students staying on course or dropping out.

The solution:

Data-driven analysis of student behavior provided the university administrators with an

understanding of the factors contributing to retention. The predictive analytical model

made connections across datasets, including those that were not readily apparent. The

university could spot, in real-time, anomalies in students’ behavior indicating whether

they are at risk of dropping out. Administrators can then quickly build intervention

strategies, including better course designs, to keep those students in the program and

on target to graduation.

Real business results:

Boosted student retention and academic excellence.

Predicted with approximately 80% certainty whether a given student is going to

drop out.

Protected university’s bottom line by establishing the conditions for improved revenue

flow.


Centre Hospitalier de NiortForecasting medical demand

“Using IBM SPSS Modeler, the IBM

SPSS data mining workbench,

we now plan to forecast the medical

activity levels of selected hospital

departments over the period 2008 to

2012.”

—

Delphine Yzèbe, of the Niort DIM

The need:

To measure annual medical activity and trends. To produce

projections of medical activity for selected departments.

The solution:

The Centre Hospitalier de Niort uses IBM/SPSS Statistics to

centralize all medical practice data and cross-reference data items on

the basis of time parameters. IBM SPSS Modeler, the IBM SPSS data

mining workbench, will enable the hospital to run activity forecasts for

future years

Real business benefits:

• A user-friendly, intuitive, simple-to-use solution requiring no

programming

• Immediately operational

• Computing power and depth of detail in processing outcomes

Attentiveness and availability of local support from IBM/SPSS

customer insight and prediction for b2b marketing 22nd ... · •revenue in the test groups is up...

Documents