Most people’s associations with technology come from media narratives about Silicon Valley. Teams of software engineers working at an accelerated pace, pumping out new software, new features, new designs. Museums on the other hand, work at an entirely different cadence. As the cost of technical labor skyrocketed in the 2010s, museum budgets failed to keep up. The result is an increasing gap between the public’s expectations of user experience and what a museum could reasonably provide. However, in the age of AI enabled software engineering, it is museums that perhaps have the most to gain.
As technology evolves and modernizes but the tech powering museums ossifies, it creates friction for the user. Friction buying tickets, friction browsing the collection, friction understanding the exhibitions. My work has largely focused on modernizing museums by identifying and resolving this friction, and I’ve generally seen tech capabilities break down into three categories: enterprise software, agency work, and in-house staff.
Enterprise software essential to a museum’s business like ticketing systems and collection management that is managed by vendors proceeds at a glacial pace. These vendors usually have such a deep lock in with their customers that expectations of new features that work for your particular museum are nearly non-existent. These ancient vendors are routinely hacked and can still rely on their customers not bailing on them because of how difficult they’ve made it to migrate to other platforms. The schemas for these products can be mind boggling and has often evolved over decades of upgrades, deprecations and changes in technology and industry best practices.
In-house technical capabilities are too expensive for most museums to consider for full time roles. This means that the top tier museums like The Met, MoMA, Art Institute Chicago, and the Cleveland Museum of Art have robust teams while more provincial organizations rely purely on agency work to create digital projects. Unsurprisingly, it is the museums that are able to prioritize in-house technical roles that feel the most relevant to our times. Cleveland Museum already has a long history of working with AI, the MoMA has perhaps the most frictionless ticketing experience of any major museum, the Art Institute Chicago is constantly innovating, and the Met has a vast archive of research available for public consumption.
Projects outsourced to agencies tend to be incredibly expensive and usually used for ephemeral projects that rarely last longer than a couple of years. The goal of these projects is more often to get press than to provide a deep lasting experience for visitors. For example, when SFMOMA reopened in 2016 there was a huge emphasis on the technological advances included in the reopening but only a few years later nearly all of the technology from the opening had been dismantled.
While agencies are always excited to work with museums as a break from their corporate clients, their interests are fundamentally different than a museum’s. Agencies bill by the hour or by the project and tend to have strong opinions about what technology they will build with. In museums without a technical overseer, I have routinely seen agencies create projects that only they can maintain, thus locking the museum into expensive maintenance if they want the project to live beyond a few months. Working with an agency can often be infuriating, because what you thought was a simple task ends up requiring many iterations to get right, this is simply because the agency is not “in it” the way museum staff are. And these frustrations largely arise when a museum’s lack of communication comes into contact with an agency’s lack of imagination. I have spent many excruciating hours grinding out very specifically worded emails about how something as banal as a click event should work. It turns out I was pretty much prompt engineering this whole time.
As AI enabled software engineering matures, enterprise software like collections systems can be tailor fit to an organization’s specific needs with 95% less overhead than before, while maintaining best practices and extending interoperability with other business needs. A savvy museum could relieve a lot of heartache with a well managed in-house software engineer while saving hundreds of thousands of dollars at the same time. If your museum is too resource starved to research and iterate for long periods of time on advancing user experience, you take inspiration from those who do have vast resources. Currently, we are seeing formerly die hard devotees of enterprise software abandon their loyalties almost overnight in favor of AI built software more tailored to their needs with far less bloat.
Today, it is not so far-fetched to imagine a museum developing their own ticketing platform, collection system, and digital asset management system while adhering to best practices and data interoperability. Rather than buy into an atrophied vendor or jump ship to an uncertain new company, a museum could just as easily build its own infrastructure. Advances in user experience rarely manifests in a shiny new technology but rather a reapplication of the everyday ambient technology in our daily lives. Software engineering is now more than ever in a state accessible to advance museum interests in both business and visitor experience.