In 2000, many colleges and universities began to digitize the slide libraries each had built for teaching art history. At the time, I was the financial and administrative officer at the Andrew W. Mellon Foundation. Knowing of the Mellon Foundation’s long-standing interest in the arts and education, these academic institutions flooded us with funding requests. The foundation was interested in helping, but it faced the prospect of writing separate and significant checks to hundreds of institutions so that each could digitize hundreds of thousands of slides (with a lot of overlap), catalog those images, ask lawyers to evaluate the intellectual property issues, and have local programmers build a repository.
We decided instead to create a new organization that could work both with content owners (museums, artists, and photographers) and educational users to solve the problem collectively—in an efficient and straightforward way. The shared collection would be far more extensive than any one institution could create, and although we would charge access fees, each institution would spend far less than it would by creating its own solution. Mellon created and provided start-up funding for ARTstor, and a few colleagues and I left the world of philanthropy to run the new organization.
We learned quickly that it isn’t easy to bring change, even what seems like exceedingly rational change, to nonprofit institutions.
Providing a service to nonprofit organizations means entering a complicated world where “mission-driven” doesn’t mean just the slogan on the institution’s Web site. Mission-driven organizations endure because society wants them to be conservative—to preserve and pass along our knowledge and values. To a certain degree, resisting change is a part of that mission.
And resist they do. When ARTstor was getting under way, I was told by one college staffer: “We have ARTstor and a lot of people are using it. But [one important professor] doesn’t like it, so we need to build our own infrastructure and system that does basically the same things for him.” In such an environment, mission determines resource allocation—not vice versa. Innovation can be resisted if it challenges anything deemed mission-fulfilling.
In other sectors, this kind of resistance to innovation can lead, as Harvard Business School professor Clayton Christensen has shown, to the downfall of established firms due to new entrants’ “disruptive innovation.” (Think of what Toyota did to General Motors or how Charles Schwab & Co. outflanked old-fashioned stock brokerage firms.) But what if society wants great institutions like the University of California, Los Angeles, or Carleton College to be sustained and strengthened rather than disrupted?
One of ARTstor’s goals was to help colleges and universities save money. But these institutions think first about their mission and what it’ll take to serve their brilliant, passionate, and some-times uncompromising purveyors of that mission. Colleges and universities teach Sanskrit or observe distant nebulae not because the market demands that they do so, but because mission-driven institutions exist to do these things despite what the market says. ARTstor’s goal had to be to help these places fulfill their mission. If we could do that, then maybe they would save resources on their own.
Today, six years after we launched ARTstor, we provide 1,300 colleges, universities, museums, and schools with software, services, and more than 1 million images gathered from about 200 collections. What have we learned so far about introducing innovation into mission-driven institutions?
Setting Up Your Organization
The first thing we learned is that even if you provide cost-efficient solutions for nonprofits, you need to aim your proposition toward accomplishing the institution’s programmatic goals. To make that case, your end goal must represent a distinctly better solution than the local alternative.
At the beginning, when some change-resistant faculty saw ARTstor’s solutions as inferior to their own campus-based solutions, we still knew we had the right idea, because faculty members’ appetite for images was literally endless. Art is being created, discovered, and photographed in new ways every day (from new angles to different light)—and it was clear that no one institution could amass all that was needed. An aggregated and networked resource, if executed properly, could do better.
We also thought that taking on a centralized solution for a thousand institutions was more effective than “letting a thousand flowers bloom.” For this particular challenge—an endlessly expanding pool of content fraught with intellectual property constraints and requiring high-quality and hard-to-find content—a new organization could create a more effective solution than could a multitasking and under-resourced local staff. Unlike the technical support staff at a university that works on many different projects, an organization like ARTstor focuses on one specific problem and employs people whose skills are geared to those particular tasks. Making digital images available for educational use is all we do.
Finally, we knew that it was important to establish that our motives and values were aligned with those of the people we were serving. Although we didn’t earn trust simply because we were a non-profit, that status helped. The question of why we were attempting this enormous undertaking wasn’t called into question the way it might have been if we had a profit motive. We continually remind institutional partners: “Your users are our users. Anything that’s good for you is good for us, and vice versa.”
Approaches to Collaborative Innovation
Many for-profit firms can mandate workflow changes from the top down. But when you want to enlist individuals who have their own strong views about the institution’s mission, it’s important to understand their work. This means understanding the staff who will need to adopt and support your solution. Staff are rightfully skeptical of an external service provider who comes in with the message “Stand back, we’re here to solve your problem.” With that approach you offer little respect for what they know about meeting local needs and exclude them from taking part in solutions.
Understanding the staff’s work depends on knowing and respecting the work of the people they support. Early on, one enraged art historian showed me a search in ARTstor for “Hagia Sophia.” More than 100 results popped up, and yet he gestured at the screen, frustrated. “There’s no view looking back at the mosque from Topkapi Palace at dusk.” He wasn’t being difficult; he needed that image to make his argument. We had to either obtain that image or enable him to add his version of it to our platform. His needs were particular, but they were real, and helping him do his work was what we had to be about. We also learned that solving complex problems takes time. And although it’s important to compromise, a compromise shouldn’t be the stopping place. Mission-driven users with high standards will give up on you if you fail to reach beyond their expectations. ARTstor’s first project was digitizing a slide library of 180,000 images. This covered the “greatest hits” and provided broad cover-age of many topics, but there were still flaws. The content coverage was insufficient for user needs, some slides were of poor quality, and others were duplicated. During subsequent years, we clustered the duplicate images, gave users the ability to use their own images alongside ours, and supplanted many third-generation derivative images, such as a slide of the Mona Lisa, with high-resolution images that allow users to zoom in on details so closely they can see the cracks in the paint. Our early users were willing to settle for less than what they wanted, and we earned some trust by responding to their needs and working to improve on our original offering.
Another lesson we learned: Even though we had a compelling vision and introduced a new approach, the solutions we provided had to be real-world solutions. ARTstor created proprietary software that allowed teachers to use very large image files within the constraints of our content providers’ intellectual property requirements. By developing the software explicitly for image users, we could add functions (zooming in on an image) that other software built for marketing presentations (such as PowerPoint) lacked.
I remember someone saying to me, “Why would anyone prefer PowerPoint when your viewer is so much better?” But PowerPoint is ubiquitous, and only when we found ways of allowing our content to be used in PowerPoint did we become a helpful part of solving users’ real-world problems.
MIT economist Eric von Hippel has observed that one reason people want to solve their own problems is that they trust their colleagues and lack faith in the motives of outside agents. Skepticism about whether an external firm’s aims align with one’s own isn’t easily overcome. Yet collaborative innovation can work. Six years ago, experienced staff at Harvard University didn’t hesitate to tell us all the things about ARTstor that didn’t work for them or their users. But today, Harvard is investing both staff time and funding, and working alongside ARTstor to build out new parts of a shared software platform.
It took time, but we learned that you can’t win over mission-driven partners by telling them where they are wrong; you can only help them do what they know is right. By respecting the work of the people in great educational institutions, we are giving them fewer reasons to try to do everything on their own. The more we succeed, the more they will, too.
James Shulman is the president of ARTstor. Previously, he was the financial and administrative officer at the Andrew W. Mellon Foundation, where he coauthored the book The Game of Life: College Sports and Educational Values.