12 Feb Software innovation helps harness brainstorming
MADISON – One success of the University of Wisconsin-Madison Reaccreditation Project gave rise to an even bigger challenge: Namely, what to make of tens of thousands of open-ended observations about the university’s future?
One clever software solution, developed by UW-Madison math graduate student Erik Andrejko, may end up finding a useful home with any organization in the throes of strategic planning. Andrejko created “Themeseekr” – http://themeseekr.com – a now commercially available service that helps collect, categorize and analyze qualitative data from surveys.
And while Andrejko built Themeseekr specifically for the reaccreditation challenge, word-of-mouth marketing led him to a half-dozen new clients, including a larger-scale current project for the Wisconsin Association of School Boards.
Themeseekr applies some of the group tagging and “crowd-sourcing” techniques common in social media software to give users a new way to harness the value of open-ended questions in surveys. But rather than building a data-mining tool, which would analyze and spit out common themes in an automated way, Andrejko says Themeseekr’s true innovation is that it functions within the brainstorming process.
“I think of this as public-engagement software,” Andrejko says. “As people participate in the collaborative process, they get unique insights into the data, rather than a computer telling you what the data says. A lot of things would be lost without the human interaction with this product.”
Andrejko’s link to the reaccreditation efforts came from his wife, Mathilde, who is a staff member of the project. Like the rest of the team, Mathilde was struggling with how to make the best possible use of the extraordinary feedback collected through the project. In total, the project generated 18,767 unique responses from students, faculty, staff, alumni and the community, around the questions of UW-Madison’s strengths, weaknesses and priorities for the future.
Erik, a computer science degree-holder who worked in the software industry, spent a rainy afternoon experimenting with a basic design that might help the cause. It led to a series of refinements and additions, and finally six distinct tools were embedded in the software that gives users different points of entry to the challenge.
Here’s how it works: Community responses from existing surveys can be imported into the software program, or users can design an e-mail questionnaire within Themeseekr itself. Then, during a team-meeting process, individuals choose a list of themes and the software program helps assign strength to those themes based on examples within the survey. It helps analyze themes across each individual in the group, showing the places where the team substantially agrees or disagrees.
Andrejko says one of the great values of qualitative surveys, as opposed to closed questions, is the potential for finding “diamonds in the rough” directly from constituents or customers. Since the program doesn’t self-select the “most relevant” responses, all potential responses have a chance to bubble to the surface during brainstorming.
“My mantra for this program is ‘don’t survey – jusk ask,'” Andrejko says.
For the reaccreditation team, Themeseekr had added value in comparing responses across different groups, showing how alumni or public values of the institution may differ slightly from those of current students or staff. For example, campus life was the top concern of current students, funding issues were the top concern of faculty, and access was the top concern of domestic alumni.
So far, Themeseekr users have been from the university and public sector. The Nelson Institute for Environmental Studies, for example, used it last year for its strategic planning. But Andrejko imagines some commercial applications that could be especially ripe. A hotel or restaurant chain that gives out customer feedback cards could add those comments into Themeseekr from hundreds of locations, providing managers with a better sense of strengths and problem areas.