Knight Lab at Northwestern: Q&A with Professor Joe Germuska
Northwestern University Knight Lab builds software that helps inform and engage the public. Their projects range from professionally developed products used by newsrooms around the world to experimental prototypes that predict voting behaviour, recommend content, and track political rhetoric.
“Knight Lab was founded in 2011 with a generous grant from the Knight Foundation. However, the actual work that Knight Lab does today has diverged considerably from some of the original intentions,” says Joe Germuska.
As a Lab, not every idea grows into a full-blown product. Some ideas are born, built, and die before being released to the public. Others move from a proof-of-concept built by students in a Northwestern classroom to the Lab for professional development and design and are then used by media makers around the world.
Is there a specific challenge or thematic area that your projects focus on?
We try to remain open to pursuing various faculty and staff research areas, so we don’t have a strict list of times or challenges. I think in short, we are interested in media design for the 21st century, taking into account new, ever cheaper devices for consuming, producing and distributing media. By design we mean best practices for successful storytelling and clear communication, adapting principles which are well understood to these new forms.
A little more specifically, we have people who are very interested in the future of visual storytelling, including Virtual Reality, Drone photography, and other advanced cameras and in platform-specific best practices for new forms such as Snapchat, Instagram, and Amazon Echo. One of my personal project areas is in bridging the gap between “open” data and “usable” data. Of course we also have a number of very popular tools for storytellers, and we are always interested in creating more, when we find a promising path.
How many people work inside?
Knight Lab has four founding faculty members, two from Engineering and two from Journalism. We also have two closely affiliated Journalism professors who have gotten involved with us since the original grant. I am not on the University faculty, but instead have a staff appointment. After me, there are three full time software developer positions and a communications manager. We also work with a number of students: we have about 12 undergraduate student fellows, and loose relationships with 3-4 graduate students. Also, this quarter, there is an academic class of five students who are physically working in the lab some of the time, and who are in some ways part of the community.
I see you have funding from the Knight Foundation, the National Science Foundation and the Robert McCormick Foundation. Do you work with any other external forms of financing, for the lab or for specific projects?
Up to now, we have not secured project-specific funding. It’s certainly something we’re open to, but it hasn’t happened yet.
What types of clients do you work with, generally?
Our work is not exactly client-driven, at least compared to my experience in other software development environments. We’ve done less partnering than was suggested in the original grant announcement, but we are renewing our efforts to find good media and technology partners.
Do you work with any news desk? If so, how close to them would you say is the process?
This has not happened much so far, but as mentioned above, we are beginning to experiment with media/content partners. It’s too soon to really characterize the process, but I will say that one of the reasons that I think the lab stepped away from the first approach to partnering was that they weren’t close enough — in the software development world I came from, committed stakeholders are critical to success, but often times, people look at Universities as a source of free labor instead of a true partner.
After the team releases a product/project, do you discuss or follow how was its adoption by the newsroom or audience?
We monitor Google Analytics for basic usage, and we do some surveys, and we also sometimes analyze data created by people as they use our tools for more insight into how they are using it and which features are more or less popular, or more or less well-understood.
How would you describe the innovation process and methods inside the lab?
We try to communicate a lot with each other about what’s interesting and what’s possible. We frequently experiment with new forms for getting that creative exchange, like community lunches with outside guests, occasional sessions where we deliver brief “lightning talks” to each other, and other ways of getting feedback and sharing information. We are also looking to some possible new models of linking work that happens in the lab with other academic work, so that students would work on lab-directed projects for course credit—the model we have in mind is very much inspired by how engineering research labs work, but we need to make adaptations all the time!