This is Ray Poynter's Typepad Profile.
Join Typepad and start following Ray Poynter's activity
Join Now!
Already a member? Sign In
Ray Poynter
British iconoclast, trying to help, educate, illuminate, and have fun.
Interests: market research, the future, politics, rugby, and sport in general.
Recent Activity
As well as following my progress below, please do visit my just giving donation page. Wow! What a day. The run was much harder than I had expected because of the terrain, lots of mud, long/wet grass, and plenty of hills. The run took little but under 7 hours and... Continue reading
Posted Aug 17, 2011 at Saints Way 2011
Every year I like to set myself a physical challenge and I like to utilise that challenge to raise money for charity. This year I am going to run the Saints Ways in Cornwall, from Padstow on the North coast of Cornwall to Fowey on the South Coast. We are... Continue reading
Posted Jul 10, 2011 at Saints Way 2011
Hi Graham, as I understand it, if I choose to Google I agree to it looking at what I do. The difference is that if you choose to use Yahoo! Mail not only will it open your emails and use them, it will open my emails to you and use them, without seeking any permission from me. I suspect that this may well turn out to be illegal in Europe as it is interfering with the mail, in the way that Phorm was judged to be illegal.
Hi Theo, I have put netnography into the catch-all, because of its smaller commercial relevance. Much as I love netnography (by which I take Kozinet's definition of ethnography via computer mediated communications)it can't be scaled in a way that competes in revenue terms with communities and social media monitoring. I think the distinction between social media as a medium to research the 'regular' world and as a subject for study is important, but again in terms of commercial market research the lion's share goes and probably will continue to go to using social media to understand brands and people, not to the specifics of the medium itself.
Hi Laura, you may find this blog post on community panels interesting Not everybody agrees on the difference between MROCs and Community Panels, but for me the key issue is whether the client feels they can divert a substantial part of their existing QUANT research to the community then it is a community panel. This tends to mean three things 1) the community is large enough to support quant work 2) the community is representative of a defined part of the customer base 3) The community is not so 'community' that its relationship with the brand is evolving too quickly and becoming too sensitised. I should also point out that some people (for example Diane Hessan and Tamara Barber) have taken the view that MROCs cannot be short-term, that they need time to develop the trust and relationships that make them work, however others, including a number of vendors and buyers favour short-term MROCs.
I would probably class social media context analysis as part of social media monitoring. I am not very happy with the name social media monitoring as it implies a passive mind set that waits to see what happens. The tools can equally be used to search for things, which is why the term blog mining was popular for a while, until blogs turned out to be such a small part of social media. There are also fields like psycholinguistics that look at the types of language being used in different contexts. A new umbrella name is needed, but social media monitoring appears to be the best to hand. Annie Pettit has made a strong case for the use of social media research to be used exclusively for the reading and searching of naturally occurring social media discourses, but I think that ship has sailed, social media research is in use to describe a much wider range of approaches and techniques.
Hi Colton The review I wrote when I finished reading the book is at I still feel that the book lets itself down, it could have been good, but falls very short, in my opinion. Some people agree with me, some dont't.
Hi Tdoyon, I think misleading is a little harsh? I do not feel that a comparison has to be apples-to-apples to be of interest. Please note, I say in the post that the chart does not suggest that analytics is replacing market research. What I am highlighting in the chart, or trying to, is that the level of interest, as measured by how often people type a word or phrase into Google's search engine, in the words "market research" is declining (and I highlight that is a typical pattern for a mature phrase). However, interest in another term used in the knowledge area, one that has areas of overlap with market research, is showing strong growth in interest (I also highlight what some of the drivers of analytics are). Another valid issue about the chart is that because Analytics is a single word it is easier for it to get more hits than a phrase like "market research". However, that does not explain the different trajectories of the curves, which is based on the interest amongst people who use Google's search engine. When I ran the Google Insights chart, most of the top news terms that it reported were about Business Knowledge and Information - confirming my view that they were interesting to compare. A few years ago ESOMAR declared that market researchers should seek to 'own' Business Knowledge and Information - a goal that is probably further away now than ever. I agree with the point that the Y axis could/should have been labelled. However, I disagree with the insult about misleading since the note at the foot of the post explains what the Y axis measures. The reason for this is that Google Insights is pretty loose in how it describes the values it produces. What they say is "Insights for Search aims to provide insights into broad search patterns. Several approximations are used to compute these results. The Insights for Search map is intended for general analysis of volume patterns. Borders are an approximation and may not be accurate.". The numbers the Y axis are an arbitrary scale based on relativities, scaled and normalised (by Google) to fit into the range 0 to 100. If a more popular item had been added, the values for Analytics would be lower, if Analytics had been removed the numbers for Market research would be higher, but the shape of the curve would remain broadly the same. I accept that I ought not to have assumed that people were familiar with Google Insights and the way it reports information. I do not accept that the intention or likely effect of the chart is to mislead.
Ah Annie, but that goes to the ontological heart of the issue. Those with a constructionist point of view would look at terms like 'validity' and 'reliability' and dismiss them as the illusions of the positivist descendants of the Vienna Circle. One of the reasons, IMHO, that online qual did not take off ten years ago (amongst most qualitative researchers) was the nature of the narrative. If you look back at the old papers which conducted side-by-side studies into online and offline, they often talked about the quantity of the verbatims, the numbers of words, the accuracy and speed of the transcripts. These 'fact' bases arguments are all artefacts with great significance to the positivists, but with little resonance to those with a constructionist epistemology. If qual and quant are to meet and understand each other, then each will need to understand the framework of the other's thinking better. Telling an artist that a bigger brush will let him paint faster will not convince the artist. Similarly, telling Ford that hand painted cars would be more personal is not likely to turn the business around.
Ray Poynter is now following The Typepad Team
Mar 15, 2010
Stop asking closed questions, with closed questions the aritmetic is easy, but you have no idea whether the questions are the right question. Ray
1 reply
Hi Emiel I think that we are looking at the end of the road for 'representative' sampling. In many markets CATI is the last techniques left that claims to be capable of 'random' samples (for example US and Australia). With the decline of landlines (not to mention do not call, call screening, shifts in working patterns), there is no way that CATI and RDD can claim to be representative. We will see a growth mobile phone panels, with voice as the main data collection method, but this is more expenive than Internet based data collection, so I do not see it being enormous. Similarly, we will see a growth in mobile phone data collection, using Java, web, and SMS, but all of these tend to be slightly more expensive than Internet, so unlikely to be massive. Of course, over time more and more people will do conventional online surveys via mobile devices, and designers of reseasrch will need to keep this in mind when creating surveys. So I do not think there is a cotradiction in the decline of landlines, the growth of mobile phones, and only limited growth for mobile phone based research. On the other hand, ten years is a long time!