I recently interviewed Tom H.C. Anderson, President of Anderson Analytics and author of industry-leading market research blog More Than Market Research. Tom is among the most active market researchers in social media. His firm specializes in data mining in combination with custom market research.
Dana Stanley: Hey Tom, how are you?
Tom Anderson: Good, how are you?
DS: Good, thanks for taking the time to participate in this podcast.
TA: My pleasure, thanks for having me.
DS: Sure. Well I wanted to have some time with you because you’re doing a lot of different things that I think are interesting to people out there in the research industry, and you’ve innovated in a number of ways. I wanted to get some time to ask you about that and let people consider some of the things that you’ve done.
TA: Great.
DS: So, Let me start out by asking you a little bit about your background and your company, Anderson Analytics. It probably would be helpful for you to start out just kind of talking about how you got into research and talk a little bit about what Anderson Analytics does.
TA: Sure, I’ve been in market research for about 14 years or so. I started out with an internship at ACNielsen/BASES and then moved on to NFO, and in 2000, I had a chance to work for a dot com, so that’s where I first started getting into social media. Then, more recently, when I went back to grad school again I got into Data Mining and specifically Text Mining, and that’s why in 2005 I started Anderson Analytics as an online research firm: surveys, online discussion board focus groups, but also leveraging text information.
DS: What kind of academic study did you do in your graduate work?
TA: I did a Masters in Economics in Sweden, and there was an area studies component as well so I did my field work in China. And then an MBA at the University of Connecticut in Marketing Research and International Business.
DS: How did you get interested in Data Mining?
TA: I had a professor who was interested in the leading edge of Text Mining and Data Mining, and I, started working with him, and also SPSS. After I was done I managed the Starwood account on behalf of TNS. And obviously, they had a lot of open ends, doing one million completes per year. Starwood wasn’t able to do much with the open ends at the corporate level, so I saw Text Mining as an obvious way to leverage that.
DS: So I think there are a lot of people out there that you know, have a general concept of what Data Mining is, but they really don’t know the first thing about how to do it, or I think more importantly how to think about what’s possible with it. Can you give some examples of the ways that you’ve innovated with Data Mining and some things that you’ve done for clients?
TA: Yeah, I mean it really varies, from analyzing and segmentation is often a key area obviously, in many cases. We’ve done a lot of work with CRM data. The client obviously collects a lot of data in doing business. So that’s usually a good step to start. Unfortunately, there are a lot of data silos, so trying to get marketing research to work more with direct marketing or CRM, loyalty etc., in leveraging multiple streams of data can be difficult. Obviously you can do a segmentation on just CRM data or you can do a segmentation just using survey data, but we’ve found the most powerful ones are the ones where we’re able to combine both of those. But for Data Mining you know, sometimes there are specific key objectives that you go in there with, and other times, it’s an exploratory to find out what trends you can find. Data Mining is discovering patterns and so, when clients are able to or are willing to go in there, and this is the case with Text Data as well, there are certain things that we’re looking for, usually, and then there are other things that we discover that we weren’t really planning on.
DS: So how do you approach a set of data, you know, without getting too specific or certainly into anything proprietary, but how do you approach, essentially a mound of data when you come across it?
TA: Well, it again depends a lot on what kind of data it is and where it’s coming from and what the objectives are, and also whether it’s just quantitative data or text data it’s also rather different. So for Text Mining, which we’ve been doing more and more of, we find that text data can come from a number of sources. Survey open ends are an obvious one, screen scraping, or blog mining etc. is another one. But we’ve analyzed, call center log data, emails, as well. We’ve set up some best practices, in dealing with Text Mining, and part of that is the AA-Text [Anderson Analytics Text], basically validation through triangulation. So we’ve used a number of Text Mining softwares, not just what we call tactical Text Mining, which looks at what’s being discussed: verbatim concepts and sentiment, but also emotional content analysis software which looks at the words people choose to use, and that tells us something about their state of mind and their emotional needs.
DS: Well, that’s interesting. With the proliferation of web sites and companies not just having one web site, but having multiple sites, I imagine that’s a big source of data for you as well. You mentioned screen scraping?
TA: Yes, often a client will ask, ‘can you scrape the web for us and tell us what they’re saying about our brand?’ And I have to explain that nobody can do that, nobody can get everything. Facebook, for instance, is very valuable, they have a host of the more personal web. Google has everything that’s public, and Facebook has our personal conversations. But that’s [Facebook] a walled garden, so most of that data, or pretty much all of it, is not available for screen scraping. So in our projects, we tell our clients that it’s much better to carefully scope out and understand what you’re going after, which sites are important. In many cases you’d be surprised at the kinds of products that are being discussed and have discussion boards dedicated to them. Everything from knitting to guitars to obviously video games and cell phones. But in those cases where the client has a product that is not being discussed as much, then I advise them to, why not look at sites where their target demographic is active, screen scrape that and and understand what their customers are discussing, and what makes them tick, so that you can then incorporate that into your marketing?
DS: OK, so lets see, so you are quite active in social media, you’re involved in a number of things relating to the research industry, on Linkedin, and Twitter and whatnot, so I want to ask you about your approach to social media and some of your thoughts of how the research industry can benefit from participating in it.
TA: Yeah, I think there’s a big opportunity – in a study that Anderson Analytics does every year among senior level marketing executives of MENG, the Marketing Executives Networking Group – social media and social networking, blogging etc. came to the top this year in terms of buzz word trends they were frustrated and tired of hearing about. On the other hand though, we found that they also felt that they were very important, so the frustration lies in not understanding how to leverage that. I think the first step in understanding how to leverage it, is to understand which of your customers use it and how they use it and how. You really need to understand social media and social networking in order to understand how to best leverage it for research as well. I view social media, Twitter, Blogging, Facebook etc…as concentric circles. You know there’s a certain percentage of the population that touches on all of those forums and then there are others that you reach on just one of those. I was an early adopter of LinkedIn, for personal business use, and from there on started using Facebook, and then started a blog, basically because we were doing blog mining and I wanted to understand it from the blogger’s perspective. I really had no, intention of blogging, I wasn’t a blog fan at all. Didn’t read any blogs at the time really, I didn’t think I could blog. I started blogging a bit, and quickly noticed that we got four times the amount of traffic to the blog as we did to our web site, which had been around for, several years, and had advertising etc. From there on, Twitter was a logical move because Twitter’s great for driving traffic to blogs. So I’ve talked to a lot of new media gurus like Seth Godin, and Guy Kawasaki, and adopted a lot of what they think, such as sharing information more openly. Not worrying so much about what’s proprietary etc., and Guy Kawasaki in terms of how he uses Twitter.
DS: What would you advise people who, are not early adopters, but who maybe have some of that sinking feeling that you mentioned the marketing executives having earlier, that this is something important, but I simply don’t have the time, maybe people are a little intimidated. What would you advise people to do to kind of maybe babystep in and get involved?
TA: Well, I think the first step is just getting on one or two of these networks. I mean, for business Linkedin is critical. So I think that’s the best step, and then you know, for personal use, maybe Facebook. But you really start understanding it once you start using it. I’ve talked to social media managers at companies that haven’t, tried too many of these social networks. Whether you believe in a certain social network or not – not that you should try all of them, but I think you have an obligation if you are working in marketing, to at least give the three or four major ones a try.
DS: I suspect there are a lot of researchers who are increasingly being asked about these things or being asked to measure these things in their research.
TA: And I think at the, at the strategic level, a lot of executives think Facebook is something their children use, and that’s not something that adults or their customers use. Either they haven’t seen the statistics, with one in three Americans using social networks and 60% of the online population, or they haven’t obviously gotten on there, and they don’t know that the fastest growing demographic, on Facebook recently has been women over the age of 35. Often children or daughters inviting them at first, ‘friending’ them, and then maybe deciding to ‘unfriend’ them, and then mom says ‘no, no, you’re keeping me on there so I can see what you’re talking about’. But also, beyond that, obviously boomers and the WW2 generation is getting on Facebook as well to keep in touch with their family and share pictures and so forth. Business networking, I think is moving a bit away from not just Linkedin, but is now also happening on Facebook. It’s nice to see the personal side of people you’re meeting with in business as well.
DS: Yeah.
TA: And I think there was some apprehension to that, do people need to see what I do in my personal life? But I think that’s evaporating. Huge social changes are coming. One thing we see for instance in our work with college students is that, whereas Linkedin tends to skew more male, Facebook skews more female and younger. So I think, in the future, women are going to be more connected, or have a better networks than men and that’s obviously a change from, from the current status, which you can see by looking at Linkedin’s demographics.
DS: Interesting, I think Facebook is an interesting case because I think a lot of people are confused about, well not confused, but maybe have mixed feelings about, as you mentioned mixing the personal and the business, but I have noticed an increasing number of people, I myself have more work contacts on Facebook, but I do have to say that whenever I put a link to my blog post up there, my mom is the first one to, to ‘like’ it.
TA: Yes…(chuckles) well, I mean, everybody has different viewpoints on social media, on social networks and different strategies, and who’s to say which strategy is right, but I’ve noticed my strategy on each of the networks has changed over time. How I used Linkedin, who I connected to and didn’t and Facebook as well. And you can see that probably by the number of connections people have on the various networks. But, there is no right or wrong way per se.
DS: Yeah, yeah, absolutely. I want to ask you, you mentioned Linkedin several times and I did want to ask you about NextGen Market Research, which is your group on Linkedin. Tell us about that, and what’s behind it, what the philosophy is behind it, and what kind of things people can expect if they check it out?
TA: Well, a lot of the groups on Linkedin, and there are obviously many, and a few in market research. For most of these it has been pretty much all about growing the total number of members for each group. Pretty much from the outset, I thought it would be nice to be a bit more careful about who got into the NextGen Market Research group. So at first it was for US professionals who had 7+ years experience in marketing research, and I ended up easing up on those restrictions, but what I didn’t ease up on was trying to keep it about serious discussion. So in other words, people aren’t allowed to post promotional messages about their own company. So I have to follow that rule myself as well on the board. But because we follow these rules, there’s much more discussion in this group I think than any of the other groups. A lot of the threads have 12, or 20 or more comments, so I get a lot of positive feedback about the group. So it’s been a very interesting way to communicate with fellow market researchers. It’s something that none of the trade orgs really at the time were offering, now they all have various groups, but they aren’t as strict about moderating their groups as the NGMR group.
DS: So people can just go ahead and click on the group, then you have some sort of validation process?
TA: Yes, Now it’s invitation only, but either myself or sometimes my assistant helps me. But pretty much I try to read all the applications and if they at least fit one or the two requirements then they’ll get into the group. And there’s the board rules that are posted there, so that people know. Basically it’s best practices for discussion board groups, which have been around for quite a while, especially in the really good forums with good community. But with Linkedin, there were a lot of salespeople getting on there just trying to drive traffic and had no idea of good discussion board protocols.
DS: Interesting, where do you see all this heading in the next few years in the research industry? You mentioned the associations are getting more involved. It seems like more and more people are getting involved in social networking, and yet there’s still a long way to go. There’s a lot more, there are a lot more people who aren’t involved. What are your thoughts on the direction of, where we may be in a few years with all of this?
TA: Well, again there’s an opportunity here and hopefully marketing research will take it. I like to think of it as it’s as big as when the internet came along and marketing research was rather quick, especially in the US to migrate surveys to the online forum. We have another such opportunity now with social media and social networks to try to leverage this, if we don’t do it, then other areas, you know whatever you call it, business intelligence or other areas of marketing will do it alone their way. But I think market researchers can add tremendous value in terms of getting good actionable consumer insights. But social media does present a lot of challenges. Often market researchers still don’t have the budget to play and experiment on social networks. Other than recruiting panelists from Facebook, how do we get all that link analysis and feedback. You know, one of the ways that we’ve experimented are Facebook applications. But we realize we have to sometimes piggyback with Marketing to get the budget, so maybe they have a dual purpose. But, we follow obviously the guidelines of ESOMAR and MRA and are clear about what will be used how. In a project we did last month for instance, it was a beta test, and we needed three thousand participants within three weeks to take part in this beta program and we had a limited budget so we used social media. The client didn’t feel the need to have a representative sample, and arguably, you don’t really know how your online panelists are recruited anyway, in many cases. And since they are going to be using social media to market these products that we tested for them anyway, it seemed logical to leverage social media for that. It was extremely successful, we did use a small group of online panelists as well so that we could measure to see what the results looked like, but that’s one of many ways obviously, or two of many ways you can leverage social networks.
DS: Yeah, there are really two strands if you think about it, there’s how researchers use social media for methodology and there’s also how they use it for their own marketing and connecting with other researchers as learning. And I think that those are both undergoing an evolution.
TA: Right.
DS: So, Tom, if people want to get in touch with you, what’s the best way?
TA: Best way, is to go to our site [http://www.andersonanalytics.com]. We have contact information there. They can email [email protected]. Feel free to shoot us any questions.
DS: Excellent, well thank you very much, appreciate your time.
TA: Thank you Dana.