A Landmark paper: the first step towards big data?
02 January 2014

Every quarter, Peter Mouncey identifies a 'Landmark' paper from the IJMR archives which constitutes a milestone in the development of the sector. Here is his latest selection:  

The utility to market research of the classification of residential neighbourhoods

Ken Baker, John Bermingham and Colin McDonald, BMRB (published in the Journal of the Market Research Society, Vol.39 No. 1, January 1997).

On November 27th, JICPOPS* held a very special lunch at the IPA in London which celebrated the 35th anniversary of the first commercial application of what later became known as geodemographics. This involved the sampling for, and profiling data from, the Target Group Index (TGI), using the then new Classification of Residential Neighbourhoods (CRN) process ( based on wards and parishes, developed by Richard Webber at the Centre for Environmental Studies). 

Even at this level, the findings were remarkable. 

This was a paper presented at the 1979 MRS Annual Conference, but not published in JMRS until the second of two issues entitled ‘Milestones in market research’, celebrating the 50th Anniversary of the MRS. 

The paper was received with immense interest, as slide after slide provided totally new insights into consumer behaviour, and identified key potential applications in media, sales, product marketing and social research. 

In the future, TGI data would be used to enhance geodems products. It could be argued that this is one of the most influential, ground-breaking, papers ever presented at an MRS conference. Why? Because in introducing the concept of geodemographics as a commercial product, it kick-starting a whole new sector, marketing analytics, and one might say the initial step towards what we currently call ‘big data’. 

This is because geodemographics began the integration of data from more than one source to aid the classification and profiling of consumers – initially census data and the post code system. 

In this paper also lies what I believe was one of the greatest lost opportunities for the market research sector. 

Instead of seizing the moment and becoming a developer and vendor of geodems, the market research sector decided simply to become a user of these products, for sampling and profiling, allowing the new marketing analytics sector to flourish, starting with CACI (with the first commercial geodemographics product, ACORN), leading to the developments at CCN (now Experian) and the founding of Dunn-Humby (now owned by Tesco). 

After the 2001 Census, there were no less than 11 systems in the UK. To date, there are 8 products planned, being updated or launched since 2011 Census data became available, including the Open Data ONS/UCL system. 

Others have come and gone over the years, such as Pinpoint. Experian’s Mosaic product pioneered the international spread of such systems (we hope to publish an example from Nigeria in due course) – I spoke at the then CCN launch conferences for Australia Mosaic as long ago as 1992. This in itself is an interested field for research as the international versions have required a high degree of ingenuity, often being created in countries lacking the robust population data sources and address coding systems we rely on in the UK. 

Geodems are no longer simply census/postcode products, they use data from multi-sources, some of which are market research surveys – becoming a real example of ‘big data’, especially when used in conjunction with sophisticated databases. 

The Landmark paper describes how geodems can be used by the MR sector, and also infers the potential value of market research data to geodems. The most obvious application has been the use of consumer behaviour to enrich the description of the clusters and to bring them to life through supporting visualisation. 

But market research data has played an important role in optimising the design of the cluster systems themselves and to build market-specific products - FiNPiN, developed by Pinpoint using data from the then NOP Financial Research Survey being an example. 

Overall, the relationship between MR and geodems has proved symbiotic and provided invaluable benefits to users. Of course, we all owe a great debt to Richard Webber for developing the concept of geodems, based on his work to identify areas of deprivation in Liverpool, later extended to cover Merseyside and the Wirral, using 1971 census data and the then new Post Office post code structure. 

Not only did Webber develop ACORN at CACI, he then joined CCN and created the Mosaic product. The wheel also quickly turned full circle as these commercial products were applied in social and public sector research (e.g. Farr et al 'Tackling health inequalities using geodemographics: a social marketing approach', IJMR Vol.50 Issue 4, which won the first IJMR Collaborative Award). 

Today, most vendors produce systems which use census statistics and data from other sources, but CACI for example, have built a product built solely from non-census sources. With the future of the UK Census under the microscope at present, the post-2021 family of geodems may have to radically alter if the current basis of the census is changed to one using administrative data sources. 

So, a further important reason for selecting this particular paper now is that as we await the recommendation from ONS after the autumn 2013 consultation on the future of the census let it provide a timely reminder of just how vital it is to have robust base-line data on the population, and what’s at stake with any alternative methodology.

In reading the paper you may think: ‘so what, haven’t we always known that?’. 

If so, it underlines just how ubiquitous geodemographics have become over the past 35 years. You had to be there in 1979, as I was, to appreciate just how revolutionary the method seemed at that time. For more information on the history and current status of geodems, see Peter Sleight’s paper on a Joint Industry Committee for Population Standards, formed in 1994 to address significant anomalies in population and household data being used in marketing, advertising and market research practice. 

* Joint Industry Committee for Population Standards, formed in 1994 to address significant anomalies in population and household data being used in marketing, advertising and market research practice.

** The original was published in the Proceedings of the 1979 MRS annual conference, but this was before the time when one issue of JMRS each year was devoted to key papers from the MRS conference (that ceased when peer reviews for all IJMR papers were introduced in 2004).

Now read the full Landmark paper

>>'The utility to market research of the classification of residential neighbourhoods'

Do you agree - does this deserve the title of a 'Landmark' paper? And please, I'd like to hear your recommendations for future Landmark papers taken from the IJMR archives.