February 29th, 2008 by Nils Maier
Sten Franke, director of ethority GmbH & Co KG and advisor at the Competence Center Online Market Research of the competence site, was interviewed by the competence site and questioned about the ethority approach.
On your homepage, information-seeking users like us really get our money’s worth. Nevertheless, the user has to be somewhat familiar with a specific topic because the homepage displays a lot of information. Web 2.0 is a major issue but, moreover, there are also matters such as digital identity, digital reputation and Word of Mouth Marketing.
In which ways should companies deal with Web 2.0? Or, to put it more straightforward, what exactly are you doing and how is this concerned with market research?
Sten Franke (SF):
I will try to illustrate this plainly. Everybody can take part on Web 2.0. The internet has changed. Today it is not just the companies showing a communication which is, to a greater or lesser extend, oriented towards the same direction. In the past, companies more or less merely displayed their products on the web, just like a catalogue, only online. Subsequently, one could also order these products online; it was, however, still one-way. A dialogue takes place on web 2.0. The consumers, namely the users, give statements and invite other users to an exchange of opinions. This proceeds on web platforms such as blogs and online communities. The use of public media, and thereby the purchase habits, has changed completely. Before purchasing anything, you compare your experiences gained concerning brands and products. Consumers organize themselves in communities, e.g. the German speaking BMW community, in which over 50.000 BMW owners, and thus BMW purchasers, participate. In these communities they discuss the brand, the specific models, their experiences and their expectations. These are millions of opinions, which are more or less openly displayed. Of course, you have to forage for them, analyze, cluster and evaluate them. This is our task, our way to carry out market research. You see, the ordinary market researcher records purchaser opinions by placing a handful of people in a room which is equipped with cameras and a one-way-mirror and then asks questions. This is what you would call a focus group. But is the composition of the group representative at any time? Is it possible to rule out any influence exerted by the interviewer? Are the group’s responses always honest and true or are there answers which are socially desirable? We are familiar with these concerns and we have the solution. The opinions on the web provide the largest focus group possible and we can rule out influences from the beginning. We are just observing.
You observe what happens on the internet? How does this work? Do you launch a search engine?
Sten Franke:
First of all, we work much more accurately, that is to say, more thoroughly than ordinary search engines. Although search engines are great for the average user, they can capture only parts of all websites. Within the Deep Web, which is the total of all openly accessable legal and illegal websites, less than 1% can be located by these ordinary search engines. We, however, locate 95%. ethority is using novel semantic analysis technologies, such as gridmaster. By means of a complex algorithm, we obtain a segmentation of groups and opinions and are able to process millions of these opinions within next to no time. On one occasion, the “Tagesschau“ reported on gridmaster and gave the résumé: “Gridmaster surfs the web like a human being, only 1.000 times faster”. Well, nowadays you could add three zeros more. This is the quantitative aspect. Our multilingual research team contributes the qualitative analysis. Thus, we are able to give answers to the most diversified questions, for example: Where and in which way is my brand positioned in comparison to its competitors? How is my brand perceived and in what ways is this perception altered by my current marketing campaign? Or by the political or social climate? – I am speaking of recent events, which have an external cause. Other particular problems are, for example, which level of customer satisfaction I have achieved and how I can improve the products or processes within my company by providing information. These are questions which concern every business company, while we give answers based on real data in real-time. At ethority, many time wasters are simply omitted. We do not have selection processes lasting for weeks, until we have accumulated a hundred participants for a survey. And, what is more, there is no need to let them argue in groups over weeks and weeks. We fish thousands of their opinions out of the web and evaluate the data. This is the foundation for our manner of market research, which is genuine and fast.
All this sounds so wonderfully easy. But it cannot be all that simple? How does a typical project with ethority take course?
Sten Franke:
Well, at first I have to say: the easiest things are often the best. But it is not as if you only need a smart monitoring technology and receive a great market analysis on demand. Ethority’s true capital is the know-how of our team of analysts concerning sectors, which was built up within almost 10 years. We were able to gain extensive project experience in all pivotal sectors. Because of our analysts being multilingual, we are capable of obtaining data from every market in the world in no time and with rather little effort. But before we settle down to work, we define the types of tasks with our clients in a briefing process. This concerns analyses of brand positioning, image quality, segmentation reports, issues and customer satisfaction as well as campaign tracking. Thus, we pose the questions which need answering. Subsequently, we analyse the semantic details in a pre-study, that is, the keywords in all combinations possible as well as the channels etc. With the help of this vocabulary of concepts, our gridmaster software, statistic probabilities, Bayesian-Belief-Netoworks and FuzzyLogic, webmonitoring debuggers are developed which we call bots. In a manner of speaking, these bots continue to refine themselves, in such a way as to analyse, segment and categorise only opinions within specific target groups. The categorised data is veryfied by our research team and post-coded in case of doubt. This is the foundation for the answering of quantitative questions. Moreover, we are able to pose further questions by including the database output at hand. These issues can be tracked down by the use of additional semantic filters. During the qualitative analysis, the obtained results in the form of blog entries or online community comments are evaluated and processed by interpreting the procured opinions. However, our job is far from ending here. We do not just compile data, we can do more and we do want more.
This takes us to the matter of your segment studies and your brand monitor. In which way are they relevant? Or are they mere instruments of acquisition?
Sten Franke:
No, it is certainly more than just that. These studies are first and foremost examples for our course of action, because the issue of webmonitoring is quite unknown to many people. In this process, we try to demonstrate what is possible. In the US you use the nice word “buzz” for any kind of spontaneous statement, which we adopted at once for our brand monitor 2.0, namely bobi, which stands for brand online buzz index. bobi is an enduring omnibus survey, which pinpoints where, how, with which level of intensity and in which topical context brands are discussed on web 2.0. Consequently, bobi is the first online brand monitor which evaluates consumer opinions on the web regarding to brand image, brand publicity, brand ranking as well as the major topics, all devided by sector and target group in monthly cycles. Based on predetermined criteria, our clients receive their own evaluated brand buzz in the course of this omnibus survey. Then you are able to compare results of the total market and identify trends favoured by the target group. Furthermore, it can be found out what your competitors are doing and how this is perceived by the consumers. You learn what makes the sectors tick. As I said before, you will get an impression of our course of action by our surveys. But, of course, this varies from client to client. The reportings all look extremely different; this implies, for example, web interface and database including visual data preparation and trend charts as well as excel files and complex management reports. Our clients often ask for strategy recommendation and, therefore, we jointly work out possibilities for an implementation. In order to do this, we also visit the respective company and work on location.
What will come after web 2.0? Will the changes continue? In which direction does ethority want to go?
Sten Franke:
We are prepared for web 3.0, you just wait… No, seriously, our task is clearly formulated: to establish our methods as an add-on, an alternative to the conventional market research. Many corporate groups are, although they describe themselves as being innovative, a bit deadlocked concerning market research. They initiate their focus groups at their favoured institute and that’s it. But we are able to convince even these creatures of habit more and more, initially for a pilot project. Consequently, they realize that we detect trends of the market much quicker and more accurate than others. For the benefit of convincing we want to expand our bobi. Soon it will be possible to use it internationally in four languages. We are developing new trend-scouting tools and intend to engage more in public matters than we did before.
