Kirsten Pijpers
Marketing & Sales Coordinator
Research Services & Software
Based in the Rotterdam office
+31 10 282 3500
Follow us
SKIM/Sawtooth Event 2011 - Conference
SKIM proudly presents the selection of papers made from all valuable submissions we received. You can visit these inspiring and varied line-up of presentations on Friday October 28, 2011.
| Title: | Turbo CBC highlights and issues |
| Summary: | In July, an advanced training event and panel discussion was hosted by Sawtooth Software in the US. Seven expert panelists were invited to give short presentations and to participate in a panel discussion (Rich Johnson, Joel Huber, Bryan Orme, David Bakken, Chris Chapman, Keith Chrzan, and Kevin Karty). Bryan will highlight the most interesting topics and conclusions from the event, as well as describe the points of disagreement and open issues sthat need further research. |
| Author: | Bryan Orme |
| Company: | Sawtooth Software |
| Type: | Software Development |
| Country: | United States of America |
| Title: | Industrial preferences in B2B agricultural markets - A comparative Discreet Choice analysis for Switzerland |
| Summary: | Managing the transition from a commodity to a differentiated good is a key challenge for industrialized agricultural markets in an increasingly competitive, price-driven world. The present paper aims at combining B2B transaction determinants and consumer preferences for food attributes by looking at industrial preferences and value chains. This research was conducted using ACBC and HB. |
| Author: | Irene Bösch |
| Company: | ETH Zurich |
| Type: | Academia |
| Country: | Switzerland |
| Title: | Two new twists on driver analysis for customer satisfaction research |
| Summary: | This presentation incites the slumbering customer satisfaction research by describing two twists on conventional driver analysis and illustrates their results with case studies and comparative analyses. The first twist brings the first fully non-compensatory preference model to the world of customer satisfaction research; the second method modifies best-worst scaling to clearly delineate which changes will have a positive or negative effect. |
| Author: | Keith Chrzan |
| Company: | Maritz Research |
| Type: | Market Research |
| Country: | United States of America |
| Title: | Comparison of software packages to set up experimental designs for discrete choice experiments |
| Summary: | The research community uses various statistical software packages to generate experimental designs for discrete choice studies. The authors will compare capabilities of popular discrete choice software packages. The packages will be applied to multiple situations (market-) researchers have to deal with. D-efficiency and standard error of estimates will be observed as statistical measures of design quality. Furthermore, usability, time necessary to operate the software and license fees will be compared to include the ‘practitioners view’. |
| Author: | Till Gissing and Josef Rieder |
| Company: | Bain & Company Inc. |
| Type: | Market Research |
| Country: | Germany |
| Title: | Product line optimization based on Choice Based Conjoint analysis using a TURF approach |
| Summary: | Line Optimization is often done using TURF Analysis. It is typically based on rather simple and sometimes biased measures of preference like purchase intention collected on a rating scale. Choice Based conjoint analysis is known to provide more valid measures of preference and willingness to pay but in combination with TURF too computationally intensive. Therefore, a TURF algorithm is presented that performs a stepwise optimization to allow for creating an ordered priority list for the introduction of products. |
| Author: | Björn Höfer |
| Company: | Ipsos GmbH |
| Type: | Market Research |
| Country: | Germany |
| Title: | „Fuzzy Choice“ – supported by Sawtooth’s new Menu-based-Choice (MBC)-software-methodology |
| Summary: | In many cases in life consumers have a clear idea of planned activities or preferences and are able to give statements about them without mentioning a precise number. There is a strong methodology developed by Zadeh named “Fuzzy Logic”. Herewith you are able to transform “linguistic” facts into manageable mathematical figures and to control systems or to calculate and predict. This presentation aims to translate such “fuzzy choices” into numbers predicting frequency of purchase and total turnover. |
| Author: | Leonhard Kehl |
| Company: | Kehl Pricing Research & Consulting |
| Type: | Market Research |
| Country: | Austria |
| Title: | Latent Class Conjoint analysis - An approach to increase accuracy of scenario simulations for very sparse data sets compared to Hierarchical Bayes |
| Summary: | The use of CBC Analysis with many parameters often creates sparse data sets which increases problems of estimating utilities additionally. Hierarchical Bayes (HB) handles sparse data sets by borrowing information from the population. However in case of uncommonly high parameter numbers HB tends to overfit the data and to create unstable partworth utilities. One way to tackle this problem is the use of Latent Class Conjoint Analysis. This paper presents several studies that all contain sparse data sets but differ in number of respondents and parameters. The results of an aggregated model, HB and LCCA are compared in terms of fit, forecasting accuracy and prediction of real market data. |
| Author: | Dorina Sohns and Matthias Tien |
| Company: | Ipsos GmbH |
| Type: | Market Research |
| Country: | Germany |
| Title: | Measuring willingness to pay for a custom-built product and its features |
| Summary: | This paper demonstrates how to measure the consumer’s willingness to pay for the features of a custom-built product in a complete product configuration, through menu-based choice modeling. We will show how we design the study and analyze the results, and how the results can be beneficial to clients. |
| Author: | Carlo Borghi, Paolo Cordella, Kees van der Wagt and Gerard Loosschilder |
| Company: | SKIM |
| Type: | Market Research |
| Country: | The Netherlands |
| Title: | Optimisation of digital TV offering through Menu Based Conjoint |
| Summary: | TalkTalk, which is a UK based telecoms provider, is entering the Digital TV market through YouView. Key objective in this research is understanding which channels are of most interest to the customer and what would be the optimum price point for each channel. As channels were not to be bundled together, Menu-based Choice was applied to allow customers to pick and choose extra channels they wanted to subscribe to. |
| Author: | Nicole Huyghe and John Irwin |
| Company: | solutions-2 and Talk Talk |
| Type: | Market Research |
| Country: | Belgium and United Kingdom |
| Title: | Which are the right covariates in HB estimation? |
| Summary: | This paper summarizes key findings from presentations on using covariates in HB at the 2010 Sawtooth Software Conference and the 2011 ART Forum. As next step we perform analyses using synthetically generated data. Two types of covariates are tried: covariate of the known preference structure and a purely random covariate. We will also compare the results from Sawtooth Software CBC/HB program with the R-Code from Rossi et al. |
| Author: | Stefan Binner and Peter Kurz |
| Company: | bms marketing research and TNS Infratest Forschung GmbH |
| Type: | Market Research |
| Country: | Germany |