Turning Data Chaos into Managed Change by Phyllis Macfarlane

ASC Labs 17th August 2026

Synopsis

How did they analyse 20,140 questionnaires for the first UK Readership Study in 1928 without computers? By hiring Chartered Accountants skilled in counting! Since then, market research has worked to bring order to the “chaos” of survey data, using advances in science and technology to improve efficiency—automating sampling and weighting, simplifying data collection (CATI, CAPI, online), and reducing workloads through passive measurement and automation. However, research data is not absolute; changes in method can alter results, sometimes negatively. Throughout, the industry has acted professionally to manage and explain these shifts. Now, with AI as the latest disruption, how should we respond? What opportunities and risks does it present, and what can history teach us?

Phyllis Macfarlane

Phyllis Macfarlane has worked in market research throughout her career, managing GfK UK from 2000-2009, then spending 2010 -2019 working on the GfK Verein’s behalf, setting up and supporting market research education in Africa and China. She is a winner of the ESOMAR Excellence Award, the MRS Gold and Silver medals, and a Fellow of the MRS. The main ‘drivers’ of her current work are improving data quality, through Better Statistics CIC (a campaign group), and education. At AMSR – the Archive of Market and Social research – her focus is to build content and grow usage through promotion of the Archive for use by A-Level Students and Academics, for Modern British History, Sociology and Politics. At ESOMAR Foundation she supports research education through Scholarships for deserving students in developing countries, 
In the early part of her career she was much involved with the development of analysis systems, including Quantum and Merlin, and would list ‘weighting’ as one of her favourite subjects.

Appendix: Links to the AMSR Archive