The Market Research Event: Day One Recap

It’s the start of the second day here at IIR’s Market Research Event. I already have two additional posts I’ll write today — but wanted to catch up on yesterdays’ deluge of information.

Many of the sessions I attended were quite good. Though I think the first session I blogged about Leveraging Facebook at Taco Bell to Enhance Traditional Research stands out as the best.

So here they are, more or less in order, with the key points I think best summarize each one:

== Facebook’s Presentation ==
After Taco Bell, Facebook’s Sean Bruich gave a presentation called ‘Are Social Networks Changing the Rules of Market Research.’ This one was a bit of a bore — but there were some interesting points.

Sean feels there are three base “rules” that Facebook (and other social networks) are changing.

Rule 1) It’s hard to find the participants you need.

Well, it’s not much of a rule, but sure, it’s hard. Facebook’s solution to this is to use their research tools and services combined with their innate segmentation abilities to find participants easily. This was when the presentation started to feel like a bit of a sales pitch. Sean did convince me that Facebook does represent a large population that is very representative of the general populous. So, the people *are* there for your studies — even older members of the population.

Rule 2: Even if you find participants they won’t want to take your survey.

Uh, yah. These rules are a bit weak. However, Sean did show that Facebook users are very into taking short 1-3 question polls in large numbers.

Rule 3: Doing a study once, isn’t the same and doesn’t reduce the cost of doing it again in another market or with a different segment.

Sean used an example of scaling internationally using Facebook. I don’t think the argument holds up. To the layman’s eye it seems that it’s going to take just as long to setup a second Facebook study as it does the first (which is admittedly quicker and cheaper than most traditional methods). In fact, the example he used about putting the same poll up for international Facebook users doesn’t make sense. What about translation for instance.

At any rate. I think Facebook is an excellent research tool to replace community building internally. It’s efficient, has a vast accessible user base and it’s very cheap. By the way, I’d love to hear from a SurveyGizmo company who has done a study with Facebook and their impression of these rules.

== SouthWest: Segmentation 2.0 – Model Using a Range of Tools and Stages ==

I love SouthWest as a company. But as an airline, well, I still prefer assigned seats. It turns out however, there is an entire segment of folks just like me and this team at SouthWest identified us. They also decided that we are not their target market and wrote us off. LOL.

This presentation was excellent. The presenter Laura Sudlow, Assistant Manager of Customer Insights did a fantastic job describing how they developed a segmentation strategy at SouthWest Airlines. They were assisted by Sachs Insights (I can’t remember the name of the speaker from that company, sadly).

Laura first explained the *wrong* way to go about creating a segmentation project. The wrong way is very little planning, quite a bit of jumping right into the survey and analytics, and very little buy in from leadership. The result? No actionable data (once they had segments defined they realized they couldn’t identify them in their database) and frustrated leadership — who had plans for the data and then found it wasn’t implemented in a way they could use.

The second time they got lots of buy in and feedback from leadership and everyone who would be using the data. They were actually surprised where and how it would be used in the organization and that changed their tactics. They also spent equal time on implementation and survey design and analysis.

The key step they claim was to start with their database. They paid a company called Axiom to segment their database first, then they spent the time adding ethnographic information to each segment by interviewing people from their database to learn who they were and what made them different. They interviewed people, sent surveys, and visited these folks in their homes to learn what characteristics distinguished them.

Now we recently went through a segmentation project at SurveyGizmo, and listening to this presentation — well, it explains the lightweight results we got back. I think in the second round we do, we’ll follow SouthWest’s example.

I have a few more to write up — but this Exhibit Hall break is over. So more in a later post!

Join the Conversation