There was an under-current of discussion at the Arts Marketing Association’s great ‘Stay Curious’ conference in Birmingham this July. Kicked off by The Audience Agency’s breakfast briefing about Audience Finder and Audience Spectrum, with the former now free to UK arts and cultural organisations (previously only free to Arts Council England National Portfolio Organisations), and reinforced by the frequently repeated references to Morris Hargreaves McIntyre’s Culture Segments from the conference stage. There are pilot MHM projects in the Northeast and in Auckland, NZ and we await the results. With consultant Katy Raines also talking about segmentation, this subject is clearly rising high up the agenda.
Andrew @TicketTattle Thomas and I have wondered in the past whether in practice the application of a ‘profile’ to a customer record really makes that much difference compared with reliable factual information on the customer from their purchase behaviour. This is very much a matter of statistics. Those of us with long memories remember working with CACI in the early 1990s on what we hoped would be an Arts ACORN, combining ticketing history data with CACI’s socio-economic and demographic data. Duncan May, now at ATG, worked closely with them on building the profile. But in the end they said the results were not statistically significant and an Arts ACORN did not emerge.
Degrees of statistical significance was of course Peter Verwey’s mantra
Degrees of statistical significance was of course Peter Verwey’s mantra at the Arts Council of Great Britain about the use of the Target Group Index data at postcode sector level. As a lottery funding assessor I read quite a few marketplace analyses in business plans which were written on the basis that a small number of people in a rural catchment area were for example factually “contemporary dance attenders” when in practice such data was only a projection and unreliable in a small population at that level. One of the benefits of Audience Spectrum is to build in real attendance data from across the country to inform the accuracy of the profile – I must declare an interest as an adviser to The Audience Agency. And MHM’s Culture Segments uses some qualitative “golden questions” to get at attitudinal and motivational factors.
Just how much information did we already have on customers
At the AMA conference Andrew and I also had a few quiet conversations with honest folk who wondered why segmentation profiles were apparently so important, just what difference could it make, how should they use it, and would it in fact really make a difference for them? This took me back to 1993 and the conversations with Duncan May, Christopher Travers, John Matthews, Vivienne Moore, Jonathan Hyams and others when I was writing BOXING CLEVER. Just how much information did we already have on customers in the new computerised ticketing systems and what should we be doing with it? My late colleague Tim Roberts mused 20 years later, after two editions of FULL HOUSE, published in multiple languages, that people were still not understanding frequency of attendance or using it to segment attenders according to their behaviour, and our 15:35:50 frequency formula remained neglected (and you can refine data today much better than that simple formula).
Michael Nabarro of Spektrix has blogged about arts organisations needing to get their heads around customer data, and they and PatronBase have expanded the tools to make analysing customer behaviour and grouping people together easier. If you want to, you can directly add in segmentation profiles. Of course, going back to 1993, we knew that actually it was not the group that was important, but each and every individual customer; CRM consultant Helen Dunnett reminds everyone about the “niche of one”: they know about their relationship with your arts organisation, but do you know about your relationship with them, and do you use that data in the ticketing system to drive the tailored communications to them?
The ticketing system suppliers are mostly collaborating with the proprietary segmentation tools emerging, but in most cases these are projecting on to individual customers a profile derived from large samples. In the past when there was just ACORN and Mosaic we wanted to test out which was “better” for targeting. But Stuart Nicolle (“You can get 35 pieces of data from 7 collected in the customer transaction”) with his “Balanced Database” tool at Purple Seven demonstrated repeatedly that real data could be used to focus and sharpen marketing effort, contacting fewer people with bigger results – we know, Helen Dunnett and I helped carry out a test at Colston Hall in Bristol with him.
“just push the button” marketing
Peter Verwey joked about one day reaching the stage of “just push the button” marketing. We are not there yet, and we can watch the segmentation wars, while recommending that people should perhaps concentrate on their actual customer data for targeting until something proven to be better comes along.