This paper will cover the practical aspects of adapting the Columbia methodology for studying campaign effects on voter decision making and the Mushiness Index scale of opinion volatility, developing the Voter Choice Project methodological approach, and what has been learned so far about the effectiveness or otherwise of these experimentations. The Columbia methodology was developed by Paul Lazarsfeld and colleagues at Columbia University’s Bureau of Applied Social Research in their seminal studies of voter behaviour in Erie County (1940 Presidential election, published as The People’s Choice (1968[1944]) and Elmira (1948 Presidential election, published as Voting(1954)). The Voter Choice Project is the first Columbia methodology multi-wave panel studying voter decision-making through an Australian election campaign. A number of changes have been made to alter the Columbia methodology from 1940s United States to 2018 Australia, and from face to face interviews to online questionnaire delivery. Additionally, the Mushiness Index, a scale of issue opinion volatility developed by pollsters Yankelovich, Skelly and White around 1980, has been incorporated both in its original form, and an adapted version applied to vote intent in an attempt to measure campaign momentum. Other questions and inspiration from many election studies around the world were also utilised in the final questionnaire design. ‘Final’ is a strong word, as tweaks have been made to respond to the volatile political environment, such as incorporating a three wave study on the five by-elections to the first two months of the study. Finally, a new way of measuring preference intent, an inferred two-party preferred measure, has been developed as part of the analysis. These experimentations in methodology are driven by necessity, the lack of research in vote causality in Australia, and the increasing inaccuracy in political polling demonstrating a need for alternative approaches that may be met by a fused qualitative/quantitative approach.