‘Traditional methods like interviewing and surveys seem to prevail and have of course proven their value along the way.’, is what I stated in December last year when discussing research on Social Media. And of course these will remain a, if not the biggest, factor in (market) research, but research in different ways seems to be more and more present.

I also used Harris Interactive as an example of a company that was already starting to use Social Media research as a means of adding to traditional research. Traditional because online surveys and private communities might be ‘new’ but they’re still quite traditional in their essence. The blog ‘Research Rants’ pointed me to a new initiative by Harris Interactive where they’re asking you to connect your ‘social’ profiles to an engine that will analyze your actions in Facebook and/or Twitter.

It seems like they’re using their normal polling panel to build a new panel that uses social media activities instead of surveys to make data-points and build a profile of your opinions and behaviors. We can only guess (thus far) about what and how they will analyze you, but it’s certainly an interesting development.

What do you think, is this ‘paneled’ approach a good thing, or is there a future where researchers will analyze everyone’s tweets, which are usually public anyway? (ie: ‘The web is the panel’) How is the panel used, is it used to make a sample and then include the (pre?) analyzed communications, or is it the other way around? Both ways seem to have advantages.

As I commented on Research Rants; ‘[a] survey might be online now, but it’s still a survey (quite often with little extra options/tricks over a ‘paper’ one), maybe even an online version of one that was (is?) conducted offline as well.’. This Social Analysis could add a complete new dimension to research, but it could also turn out that in fact all these communications tell us about the same as traditional research…

Whenever people asked what it was that I meant by “doing Social Research”, I could talk passionately for an hour. But sometimes it became apparent that a more structured approach was needed. After all, speaking in a passionate way about something you have in your mind might not always get the message across in a clear way. Thus I developed a way of structuring my talks.

I think my main argument for both how Social Research is possible and why it could be useful can be summarized in three main points. I’ve used these three points from then onwards to structure my approach when talking about this issue, which seems to work fine for me now. I’ve made a slide show based on these three main points, which may be nice to share with you here.

First of all of course my three points; these are:

• Explain survey based research; why are we used to it and why are we using it. Is this type of research flawless, does it have any shortcomings?
• The future of research is not about presenting you with the crowd, it’s about presenting you with their opinions. Social research makes it possible to tap right into opinions instead of people.
• We have to set up Research and show it is not only possible, but that it works. Do not wait for demand, be the unique company that can supply this extra knowledge and, in a way, create the demand.

Of course presenting you with this slideshow without me talking might seem a bit useless, but I think you can gather some information from it anyway. And besides; I’m always willing to give ‘a talk’ about this, of course.

CC Photo at the start of this article: Matthew Bowden
The nice picture of me on the fields of HAR2009 at the end of the slideshow, was made by Anne Helmond.

Following the recent acquisitions in the research industry, think about Toluna buying Greenfield Online/Ciao Surveys, one would expect a response from other players in the market. And to be a bit more specific; what would smaller players in the market do with their panels? There are several obvious solutions to the troubles of maintaining smaller panel(s), one could thing about an advanced sharing system or outsourcing panels.

As all of these seem to be happening there is the other obvious solution to which I’ve been pointing several times in the past; several companies ‘chipping in’ together and combine their panel efforts.

And indeed; now several European research agencies have committed themselves to pooling their panels together and create a new fieldwork business called Panelteam. Founding members are Multiscope, Mo’web Research, M3 Research, Panelbase, Directpanel Research and SWG according to Research-Live. Peter-Paul Laumans, now working at Multiscope, will be heading this new company, no doubt having to navigate it to the hard times of crafting a single structure out of the separate panel entities. It will be interesting to see how fast they can set up a smooth working new organization out of these separate resources.

Even more, it will be interesting to see what kind of influence this pooling has on the market. No doubt Panelteam would like to establish itself as an important player in the field, but they will feel the competition of both existing larger multi-country players as well as others pooling together their resources in a similar way. My guess would be that a next step might well be consultancy firms combining some of their research efforts in a similar vein.

Have you ever taken a survey recently, or maybe several surveys? If so you might have wondered about several things. And maybe those things are exactly those things that bother me about surveys. One could have, of course, something against surveys altogether, but they remain a powerful research tool, that is one thing that’s for sure. Even though I’m personally calling for a broader way of doing research, I don’t think we should abandon it.

I do think we should look more into some of the things that come to mind when you look at a survey. These things are often related to a certain tradition that lies in doing survey based research and indeed in a sort of tradition that is created per tracker survey as well. Not changing things for tradition’s sake, is one of the worst things a researcher could do and still this seems to happen all too often. Survey methodology in general is one thing, but a lot of me wondering about survey research is the lack of changes, of innovation.

Tradition

If any objections about wording, questions, long surveys and tradition sound familiar to you, you may have read Jan Hofmeyr’s (Synovate) Research Magazine interview where he comes up with similar notions. He describes in an honest and clear way how tradition seems to hold the Research Industry hostage some times. An excellent example is about brand tracking; surveys in these studies seem to be way too long, which makes response and dropout rates suffer and will of course cost the researcher way too much money. Still companies seem reluctant to step forward and possibly lose half of their income and clients seem to be reluctant because the ever present danger of losing the budget for your department if you seem to be able to spend less this year! Quite a classic way of holding on to traditions for little more than some sort of business continuity reasons.

To these observations I would add that an other reason is often because one likes to compare studies. This, of course, mainly concerns tracker-like studies. Questions can sometimes be added, but clients could be very reluctant to remove any, or even improve the wording because of reasons like “this has been running like this already” or “we want to compare this with 2001”. In these cases, the question of course is whether comparing “bad” data (defective, in the sense that Jan is using it.) with new bad data makes sense at all!

Can we add to survey research?

On other point is the research itself; I mean: come on, survey research? Can’t we come up with some nice additions to that? Well, of course we can, and I believe that we can even strengthen survey based research with additions like these. Still, very little seems under way to do this, even with the “2.0” hype going on like crazy. Now of course I do see the point in no jumping on any bandwagon, but the way clients and others seem to hold to just surveys some times seem to be a bit odd, to say the least. When one looks at comments made by Hofmeyr like

“it’s very hard to get people to change. We’re supposed to be researchers; we’re supposed to care about truth and new knowledge”

and

“Our industry is one of the most conservative you will come across. Look at IT – look how keen they are to throw out the old and adopt the new.”

it’s maybe not that hard to see why a “knowledge” industry seems very reluctant to use any improvements, let alone new ways of doing things.

Being this knowledge based industry I think the ideas, the knowledge, is there, what we have to do now is to use it! This will not happen tomorrow, but if we can come up with a good way of doing “new research” tomorrow, present it next week, clients may follow and demand these things next month, if you catch my drift. I think this is a train we cannot afford to miss or else we might never arrive in the future while others who did “get it” are catching up and do get there in the end. So, of course, proceed with caution, but be sure to catch that mediatrain that will departure very soon or has perhaps left the station already.

Text-analytics on data present on the internet, in a way a sort of quantitative way of doing research that is traditionally qualitative seems to hold a lot in store for research. But despite the fact that this is actually one of my personal favorites for the future of Market Research, little seems to really take off. Sure there are talks about doing research on twitter, but very few real efforts have been made or at least way less than one would maybe expect. For now most of marketing efforts seem to regard online communities as just another advertising space, but things slowly start to change.

Harris Interactive, a company in the Market Research Industry, has started using text analysis software from Clarabridge to add an extra dimension to their traditional survey based research. For their Harris Poll they will use this tool to measure this important source of opinion. According to Research-Live, the first test was done to research opinions on Obama’s healthcare reform initiative.

Of course there is the question on how one can possibly take advantage of this huge stream of information without getting lost in the amount of data gathered (or that could possibly be gathered). Indeed the idea that it is hard to make something out of this ever growing stream of conversations seems to be one of the causes for the hesitation to adapt any of the upcoming technologies and integrate them with other research. Then there is, of course the sense of tradition that seems to lurk over research, be it market-research, social-research or any other type of research.

Traditional methods like interviewing and surveys seem to prevail and have of course proven their value along the way. Clarabridge is (of course?) sure about the added value of their technology. They emphasize the possible richer perspective this type of research can add. Their analytics should be able to both filter out any non-opinion conversational data and quantify positive and negative comments. Maybe the first breakthroughs for research on “social media” will be in 2010, after all the chatter there was about it in the beginning of 2009.

A while ago I tweeted about a weblog that criticizes some of the online research that is being done. And I would say: the author has every right to do so! Because making a survey seems easy: it certainly isn’t.

With all the research professionals that are out the one would think that things can’t be all that bad, but the weblog “Bad Research: No Biscuit” shows that there are still a lot of things that are questionable to say the least. Anonymous research professional “researchrants” describes some of his worst research experiences.

I’ve highlighted some of the problems that “Researchrants” talks about, on the Dutch online marketingblog “Bijgespijkerd”. If you can read Dutch, or are not afraid to use google translator, you can find it here.