More than 250.000 organisations worldwide use SPSS statistical software. This was a chance for delegates to see how leading bodies use SPSS and its extension modules every day to carry out clustering, modelling and dozens of other types of data analysis easily and effectively.
|
Dimensions
Going beyond market research
René A. Scherer
Head of IT Survey Competence Centre
Credit Suisse
Switzerland
Credit Suisse Group is a leading global financial services company headquartered in Zurich. As an integrated global bank, Credit Suisse provides its clients with investment banking, private banking and asset management services worldwide.
The company’s IT Survey Competence Center is a full-service provider for Enterprise Feedback Management (EFM) across Credit Suisse’s customers and employees. It started in the year 2000 with paper surveys, and is now using Dimensions to develop up to 100 different projects per year.
The surveys are primarily performed for internal customer groups (IT, HR, Banking, Finance etc), but the centre also works as a contractor for the bank’s Market Research team, targeting external customers.
The survey portfolio includes regular surveys on IT customer satisfaction, IT project satisfaction and support centre surveys; ad hoc surveys about products, services and processes; and HR topics (eg mood barometers).
This session was about the benefits of using Dimensions outside the usual context of market research to conduct real EFM within a large organisation.
Back
|
Clementine, Text Mining for Clementine
Mining Dutch history: researching public debate in the nineteenth century
Dr José de Kruif
Researcher
Utrecht University Research Institute for History
The Netherlands
The presentation discussed research into the role of pamphlet and newspaper texts in the Netherlands during the 19th Century, a period of transition in the Dutch media landscape.
Many textual sources traditionally used by historians are becoming available in electronic form, opening new possibilities for the use of OCR and text mining to examine large quantities of printed materials. The paper described how pamphlets and newspapers covering important public issues were scanned and, through OCR techniques, converted into electronic texts.
Then, using text mining, the Research Institute compared texts for common occurrences of rhetorical devices, terminology, references to persons, institutions, events and arguments.
Text mining was also used to compare the results with metadata such as the medium, authorship and genre of the publications.
The session described how this method results in a much clearer image of how the media worked in the 19th Century, and how it offers more concise answers to the questions of what messages were being conveyed, who they were meant for and how they were being communicated.
Attendees also heard how the research completed – and partly replaced – fragmented impressions of nineteenth-century press culture in the Netherlands, especially during a period of passionate dispute between the Government and Protestant and Roman Catholic citizens.
Back
|
|
Dimensions
Why perform benchmarking?
Torben Liborius
Senior analyst
Deloitte Denmark
Denmark
At this session, the Danish arm of one of the world’s leading consultancy firms proposed an answer to a simple question: why perform benchmarking? The question is not as trivial as it might seem, and the session heard how to clarify the purposes of benchmarking before going on to exercises focusing primarily on non-financial data.
Those attending this presentation learnt about the selection of parameters, how to find and collect benchmark data, and how to provide reports suited for follow-up. They also saw an example in the form of a ‘health care audit for businesses’, a tool designed to assess the perceived status of health care, stress and well being among the employees of a business.
Back
|
Clementine, Text Mining for Clementine, Dimensions
The Predictive Enterprise at work: revealing the causes of dissatisfaction through analytics
Simon Dudley
Customer Analytics Manager
Royal & SunAlliance
UK
Royal & SunAlliance is one of the world’s leading insurance companies, writing business in over 130 countries and providing general insurance products to over 20 million customers worldwide.
This presentation discussed how Royal & SunAlliance and SPSS formed a partnership to understand better the interactions that take place across their sales, service and claims call centres, and how this insight drives operational change to enhance customers’ overall experiences.
This real-world example of the Predictive Enterprise at work described the benefits that have been realised in the areas of increased customer satisfaction; improved customer retention; and optimised operational processes through understanding the root causes of customer dissatisfaction – why people phone call centres and how many calls they make before the issue is resolved.
The session also included practical advice on how to approach this type of analysis: the challenges to expect and the lessons to be learned.
Back
|
|
Clementine, Text Mining for Clementine
Credit scoring in real time: reducing risk with text mining
Piero Biagi
General Director
Nolé SpA
Italy
Nolé SpA, controlled by Banca Agrileasing, is one of the leading Italian companies in the growing field of hiring innovative, all-inclusive value-added equipment solutions. Its clients include business and professional customers in the telecommunication, information technology, copying and digital printing, medical and banking sectors. Nolé operates in partnership with a group of vendors who supply the equipment offered for rental.
In particular, Nolé is the market leader in ‘small ticket’ rentals or leasing operations valued up to €25.000.
One of the biggest challenges facing rental businesses is assessing client credit risk, and in this session Nolé described how it used data mining technology to build an efficient real-time credit scoring system, and how the addition of text mining provided better predictions. They also presented results – a 55 percent increase in new business with no additional staff, and early defaults reduced from 3,6 percent to 1,5 percent – that demonstrate the effectiveness of their approach.
Back
|
Predictive Analysis at SAP
John MacGregor
Product Management Director
Business Objects, an SAP company
UK
Michael Adam
Director, Solution Management
SAP
Germany
Traditionally, Business Intelligence (BI) has been concerned with reviewing historical data and monitoring current data. However, to truly contribute to decision making it must supply a predictive dimension. This presentation showed how SAP, Business Objects and SPSS have combined to deliver a unique solution to meet the requirements of data analysts plus, importantly, BI end users (even casual or novice BI end users).
At this session, attendees also learnt about:
- Specific examples of Predictive Analytics for banking within SAP Bank Analyzer
- The details of strategic partnership between Business Objects and SPSS.
Back
|
|
SPSS
Analytics and the art of motorcycling
Dr Alan Tilly
UK
According to the recently published Stern Report, climate change presents a serious global risk. And as motor vehicles are one of the main sources of carbon dioxide emissions, the use of more sustainable modes of transport is being encouraged.
However, the role of motorcycling as an alternative to private cars is unclear: while motorcycles are cheaper to run and offer the same door-to-door convenience, riders are particularly vulnerable to accidents. At this presentation, delegates found out how analytics were used to research the role of motorcycling in the 21st century.
Building on a literature review that identified the costs and benefits of motorcycle use, the presenter conducted a survey that generated 8.174 completed questionnaires. This was designed to construct a profile of motorcyclists, their motivations and opinions, how they used motorcycles, and the number of accidents they had. The project also examined external factors that influenced the use of motorcycles.
This session demonstrated how, provided certain conditions are met, motorcycling can play a useful role in reducing carbon dioxide emissions while offering mobility and other advantages such as reduced congestion and easier parking.
Back
|
Dimensions
Know your customers: how Braun keeps in touch
Stefan Bender
Vice President, Marketing and Strategic Research
Braun GmbH
Bernhard Witt
2x4 Ltd
Germany
‘Know your customer’ is a pretty good rule in business, but really knowing consumers’ personal opinions is one of the most difficult challenges for companies such Braun, a subsidiary of Proctor & Gamble.
At Braun, they manage this task through a strategic product registration platform that allows them to get very fast feedback after the launch of new products or a change in marketing activities.
At this session, Braun described how it obtained rich consumer profiling across most of its products and across key regions.
The presentation included descriptions of its techniques for data collection, data processing and ad hoc analysis, as well as outbound direct marketing to deepen consumer insight and/or cross-selling.
Other topics included:
- How to implement a product-registration environment using multi-language platforms
- How to establish direct relationships with consumers despite independent sales channels
- The pitfalls to avoid
- International comparisons and their differences.
Back
|
|
SPSS
A look at SPSS 16.0 and a preview of SPSS 17.0
Marcus Hearne
Manager, Product Marketing, Statistics
SPSS Inc.
USA
SPSS 16.0 continues our rich tradition of regularly adding significant new elements to our flagship software. SPSS users wanted to attend this session to learn about new SPSS 16.0 features such as:
- A new Java interface enabling Windows, Mac, and Linux versions of SPSS
- A searchable Output Viewer, resizable dialog boxes and much more
- More powerful statistics, including a new Neural Networks add-on module and enhanced algorithm support
- Improved data management features such as Unicode support, import/export of Excel 2007 data, and an enhanced Data Editor
- Better programmability through Python
- Support for the Predictive Enterprise with multi-threaded algorithms, and the SPSS Adapter for SPSS Predictive Enterprise Services.
As a bonus, this session also included a preview of things to look forward to in SPSS 17.0.
Back
|
SPSS, Clementine, Dimensions, SPSS Text Analysis for Surveys
Using analytics to develop stronger and safer communities
Keith Bentley
Chief Superintendent (Retired)
Greater Manchester Police
UK
In the UK, responsibility for delivering crime reduction and public safety initiatives, at both strategic and tactical levels, falls to the police and to local authority personnel. In Oldham, a town near Manchester, the local government authority and the police formed a strategic partnership with SPSS to integrate data collection with traditional surveys and analysis methods to help deploy community safety platforms.
People concerned with establishing and maintaining safe and secure communities were interested to hear how Oldham Police used these analytical methods to embed crime reduction and public safety improvements within local communities through ‘Local Area Agreements’ that provide a road map for sustainable improvements in public safety and securing government support.
While SPSS Base and Dimensions were the primary tools for this exercise, the session also discussed concepts for linking SPSS Clementine into an intelligence model that may be helpful to police operations.
Back
|
|
SPSS
Improving social welfare policy-making with SPSS reporting
Carlo Vreugde
Analyst Co-ordinator
SGBO
The Netherlands
This presentation particularly informative for those who use SPSS Base to analyse data, and then output the results to Microsoft® Word.
SGBO is a research and consulting agency specialising in local government issues (formerly, it was the research arm of the association of Dutch municipal governments). Following the introduction of a new law that makes local government agencies responsible for a wide range of social services, SGBO was commissioned to produce benchmark reports for more than 150 municipalities.
To achieve this efficiently, SGBO used SPSS syntax to export hundreds of different analytical results to Word documents, not as blocks of text but automatically placing text, tables and graphs in the placeholders and bookmarks set up in Word.
SGBO went on to describe how automated reporting has improved the effectiveness and cost-efficiency of local government, enabling municipalities to minimise local taxes.
Back
|
SPSS, SPSS Server, Clementine
Using SPSS to validate Basel II requirements
Giacomo Petrini
Head of Models and Process Validation Services
UBI Banca
Italy
Under the first pillar of the Basel II Accords (International Convergence of Capital Measurement and Capital Standards), banks are required to assess how much capital they need as a safeguard against various types of financial and operational risk – including credit risk.
Banks must also establish an independent internal team to validate the methods used to perform the risk assessment – a very complex and time-consuming process.
This session was about how UBI Banca, a regional Italian bank with total assets of €119 billion, used statistics and data mining to simplify and speed up validation of its implementation of the Internal Ratings-Based approach to assessing credit risk.
The presentation focused on UBI Banca’s use of predictive analytics to validate internal estimates for credit risk and to verify the IT architecture of the internal rating system, making it easier to register the validation process with Bank of Italy in a relatively short time.
Back
|
|
SPSS
Programming SPSS to control processes and drive performance
Klaus Bergmann
Manager, Methods and Product Development
Oliver Hülser
Senior Specialist
GfK AG
Germany
Germany’s largest market research company (and the fifth largest worldwide) considers itself as “a supplier of knowledge”. This is one reason why GfK focuses on innovation and progress, ensuring that its methods, tools and practices are state-of-the-art and deliver consistently high quality information.
In this session, GfK explained how the company revitalises its methodologies and how it uses SPSS’ programmability to control complex processes to improve performance and provide methodological options.
To illustrate these processes, GfK presented its approach to segmentation, data fusion and imputation, and described how they proceed from simple clustering to a complex segmentation process. The presenters also discussed how GfK takes advantage of SPSS Base’s programmability for all the methods it uses to deliver knowledge to its clients.
Back
|
Clementine
Constructing a robust definition for churn modelling high-value customers
Sarah Gray
Customer Insight Manager
Panyiotis Georgiou
Customer Insight Analyst
Tesco Mobile
UK
As offerings amongst competitors become increasingly rich, operators in the UK telecom industry are continuously faced with the issue of customers churning – defecting to other suppliers. The challenge for Tesco Mobile was to build a model to identify those high-value customers with a high propensity to churn.
This session described how Tesco Mobile built the churn propensity model, how predictive analytics was used proactively to define an approach to retaining high value customers, and how this has been deployed in targeted marketing campaigns.
The presentation also demonstrated how an iterative procedure was used to arrive at the modelling definition, and how this was a key to the model’s success. It also shared practical tips and tricks based on the project.
Tesco Mobile’s predictive model delivered the capability to apply retention strategies towards customers with a high propensity to churn, and to vary the strategies according to the customer’s value. The intelligent application of predictive modelling has helped Tesco Mobile to control churn and enhance the returns from its retention programmes.
Back
|
|
Clementine
Churn modelling for Koç-Group companies
Enis Basegmez
Analytical Business Intelligence Manager
Melisa Topcu
Business Intelligence Consultant
Tanı Pazarlama ve İletişim Hizmetleri A.Ş.
Turkey
Tanı Pazarlama ve İletişim Hizmetleri A.Ş. (Tanı) is a company within Koç Holding AŞ, one of Turkey’s leading industrial conglomerates with operations in the automotive, durable goods, food, retailing, energy, financial services, tourism, construction and IT sectors. Tanı provides Customer Relations Management (CRM) services to both external companies and to the Koç Group, in particular its customer satisfaction and loyalty programme.
As markets in Turkey mature, shifting from rapid growth to near-saturation amid fierce competition, the focus of Koç Group companies shifted from building a customer base to keeping customers ‘in house’.
In order to thrive in such market conditions, Tanı produces life-stage and event-based marketing activities; creates and manages multi-channel campaigns; determines business needs and fulfils them through data mining and other predictive analytic techniques.
Tanı also leads Koç Group’s CRM projects, including segmentation modelling, loyalty programmes, Recency-Frequency-Monetary value, migration, and churn analyses, which are all designed to attract attention and improve loyalty.
This session discussed Tanı’s methods, which include:
- A digital loyalty card platform that recognises customers and transaction details across 15 companies in seven sectors
- Pre-defining customers who are more likely to churn
- Methods to keep the customer ‘in house’
- The competitive experience, which creates cost and profitability advantages, gained through thousands of campaigns.
Attendees also heard about how Tanı’s churn model has successfully identified over 100.000 customers who were likely to churn, and placed them on the loyalty programme.
Back
|
Clementine, SPSS Predictive Enterprise Services
Faster, more flexible development of statistical processes with SPSS
Dr Helma Schapendonk-Maas
Statistical Researcher
Statistics Netherlands
The Netherlands
Like many official statistical agencies, Statistics Netherlands (SN) is facing a continuously changing environment: new, often large data sources appear, and existing ones change.
To cope with the dynamic processes required, and with changing business rules, the organisation needs tools that can be adapted for new types of data exploration and facilitate easy development.
As a result, SN used Clementine and SPSS Predictive Enterprise Services to develop new statistical processes, with the aim of achieving faster development times, more development flexibility and a more uniform process environment.
This presentation described how, using data mining, new processes were developed for demographical statistics that were not as dependent on support from IT experts and could easily be maintained by statisticians, and how Clementine proved very suitable for transforming large datasets with relatively few variables.
In addition, SN described how SPSS Predictive Enterprise Services ensures reproducible and traceable results, including version control and audit trails; how it supports iterative development using Clementine on the actual statistical data; and also how it provides a generic tool for maintaining and using statistical rules.
As a result, Statistics Netherlands developed, tested and put in production multiple streams for many different statistical products while meeting the planned development schedule. Building similar processes using standard development methods and tools would have taken two to three times longer.
Back
|
|
Dimensions, SPSS Text Analysis for Surveys
Introduction to SPSS Text Analysis for Surveys: don’t leave open-ended questions untapped
Eric Martin
Product Marketing Manager
SPSS Inc.
France
This session was designed for SPSS users who ask open-ended survey questions, but do not use SPSS Text Analysis for Surveys to categorise or ’code’ the text responses. It was also suitable for Dimensions users who are interested in automatically coding survey text responses.
In this session, attendees learnt about the objectives behind asking open-ended questions, and the challenges of categorising text responses. They also heard descriptions of the product’s features and benefits, and understood how SPSS Text Analysis for Surveys complements other SPSS products.
They saw how easy it is to import text responses, extract and classify concepts, refine extraction results, and categorise responses, while the analysis of customers’ opinions and feelings was also discussed.
Back
|
BI and Predictive Analytics: enhancing and extending the promise
Carol Martin
Director, Market Development
Cognos, an IBM company
Canada
As a result of the integration between Cognos’ powerful performance solutions and SPSS’ Predictive Analytics technology, Cognos and SPSS customers can now monitor and analyse their historical and current performances as well as forecast and predict their future performances.
In this session, Cognos talked about how to extend the value of predictive analytics to decision-makers across an enterprise by leveraging existing Cognos or SPSS investment, using the power of the combined relationship to more rapidly and confidently anticipate and respond to changes in market conditions, risks and customer behaviour.
Back
|
|
SPSS
SPSS Expert Session: using programmability and scripting to extend SPSS and increase productivity
Jon Peck
Principal Software Engineer
SPSS Inc.
USA
In this session, people learnt how to extend the scope of what can be done with SPSS to solve problems with programmability. It was an opportunity to find out about what’s new in SPSS programmability and scripting or, for those new to this feature, get an overview of programmability.
Among the topics they heard about were:
- More data management: using the Dataset class for complex data management tasks
- More statistics: using R programming language packages, including programming R graphical output in the SPSS Viewer
- User procedures: using the Extension facility to create your own SPSS syntax to run Python or R programs
- Front-end scripting: using Python for Viewer and user interface tasks, and for autoscripting.
- The role of Basic scripting
- Unicode: working with programs when SPSS is in Unicode mode
- Enhancing SPSS procedures: using programmability to extend what a procedure can do.
These were illustrated with examples, such as:
- Comparing the variable dictionary and/or data of two .sav files
- Producing a codebook with dictionary and statistics using SPSS-style syntax
- Searching and transforming patterns in string variables using regular expressions
- Customising category sorting in Custom Tables output, and adding flexibility to subtotals
- Running an R package on active SPSS data to get output in SPSS Viewer pivot tables and charts.
Back
|
Dimensions
Improving the recruitment process with predictive analytics
Dr Peter Holderegger
Organisational Psychologist
Switzerland
Helvetia is one of Switzerland’s five largest general insurance companies, with annual premiums of 5,3 billion Swiss francs (€3,4 billion). Personlisation is an important element of its sales process, and having the right people can make the difference between closing or losing a sale.
Recognising the importance of its staff’s personalities and capabilities, Helvetia used predictive analytics to develop an online tool that could pre-select job applicants. The goal was to improve hiring accuracy by setting clear criteria for potential candidates and eliminating arbitrary decision-making.
This session was about how the new tool helped cut costs and save money in the recruitment process; how managers spent less time interviewing and reviewing applications from unsuitable candidates; and how the recruitment process has become far more efficient. Attendees also heard that, as a result, the employee turnover rate has dropped from 15 to nine percent in the company’s three largest regions.
Back
|
|
Clementine
Customer loyalty programme supports business strategy at SNCF
Sébastien Le Lardic
CRM Project Manager
Aurélie Amira
CRM Project Manager
CRM Services SNCF
France
The CRM division of SNCF, the National Railway of France, provides relational marketing programmes to support the operator’s sales strategies for its high speed and regional train services.
Those attending this session discovered how data mining technology was used to analyse the traveling behaviours of three million SNCF customers, leading to the development of loyalty programmes targeted at specific groups.
In particular, they found out from two examples how data mining technology was used, first, to carry out a Recency- Frequency-Monetary value analysis to segment the customer base and reveal different classes of ‘Grand Voyagers’ (heavy travelers); and second, how predictive analytics was also used to identify those customers with a propensity to travel first class.
They also heard how specific loyalty programmes were developed for each of the customer groups.
Back
|
SPSS, SPSS Server, Clementine
Torturing your data: an analytical CRM success story
Andreas Kokkinos
Head Business Intelligence
Marfin Laiki Bank
Cyprus
Marfin Laiki Bank (Laiki) is the second largest bank on the island of Cyprus, and is growing rapidly throughout parts of Europe and Australia. However, rapid expansion in other countries caused the bank to lose focus on its core business in Cyprus, resulting in a slowdown in the growth of market share for credit cards.
After identifying the problem, Laiki launched a major effort to reach out to its customers. Using predictive analytics allied to data analysis methods, it examined areas such as churn prediction, value segmentation and campaign development.
A joint effort by the Cards, Marketing and Business Intelligence departments not only stopped the slowdown but turned it around: for the first time in its history, the bank’s rate of market share growth in credit cards is significantly higher than that of the market as a whole.
The session explained how this success was due to concentrating on basic marketing and CRM, and once more becoming familiar with the customers. It also described the processes and the functionality of the tools that were used, and how predictive analytics technology made the project much easier.
Finally, Liaki talked about the moral of its success story – “If you torture the data long enough, it will confess” – and how this approach is now being applied to all other major banking areas with quite impressive results.
Back
|