Is there an audience out there for classic natural history programming à la David Attenborough, but dubbed over with more absurd commentary from the comedy band Flight of the Conchords (“New Zealand’s 4th most popular guitar-based digi-bongo a-capella-rap-funk-comedy folk duo”)?

One possible way for a programmer to answer that question would be a long show incubation period, ironing out issues internally and with meticulously chosen focus groups. Another might be to present publicly a less polished product for a short time, and let audiences decide for themselves.

The BBC’s Taster platform is an audience-facing holding pen for new ideas across the BBC’s many divisions, whether an experimental Flight of the Conchords series, or an interactive quiz infrastructure that can be used across different BBC programs, or Rio Olympics coverage in 360° video. Pilot projects are presented on Taster for three months (the occasional project may be up for longer), during which time visitors can try, rate, and give feedback on whatever interests them.

“We have lots and lots of creative communities inside the BBC who make TV shows, radio shows, who build and create content for websites. In their day-to-day, some of those teams are reaching for the next form of storytelling,” said Will Saunders, the BBC’s creative director of digital, who manages the editorial content that goes onto the Taster platform. “The next thing they want to understand is whether or not they need to do things differently, or even add a new character, or a TV pilot. If they want to, they can test their ideas out using the platform”

What’s unusual about Taster is not, of course, the concept of piloting ideas, but the fact that the BBC was willing to show the public projects in a half-baked state, glitches included. Taster, which formally launched in at the beginning of 2015, is an audience-facing component of Connected Studio, which has been fully integrated into BBC Research & Development (if you’re interested in the technicalities of this structure, this post lays out some of that). The platform has launched over 120 different pilots, working with all the varied departments within the BBC as well as outside organizations. Since then, it’s seen 43 million pageviews and collected over 110,000 pieces of audience feedback, according to Connected Studio executive product manager Eleni Sharp, who looks after the Taster platform.

“One of the problems of being a well-established web presence is that it becomes very difficult to develop new concepts, formats, and ultimately new products,” Adrian Woolard, head of BBC’s Connected Studio and its R&D North Lab, said. “We’ve done a series of programs over the years where we’ve tried to do that within the main parts of the business, tried to get them published through BBC News, BBC Sport, and so forth. But it’s difficult to get them exposed at a sufficiently early stage. The expectation is very much around mature, universal products and services.”

There was a gap, then, in the development process of getting immediate feedback, and getting it earlier, from a larger swathe of potential viewers (readers, listeners) to decide whether an idea was further investment.

“But everybody could see that the Internet offers you the power to put up stuff early, to learn from your [minimum viable products]; it’s increasingly fast paced in terms of decision making and feedback from audiences. So that’s where Taster came from,” Woolard said. “Every project we put up had an unknown. We wanted to answer a specific question, or for some of them, we wanted to answer a couple of questions. Everything could be published and trialed with an audience that could then inform next decisions.”

When visitors try out a project on Taster and then hit “rate,” they are guided through a short multiple-choice survey to ascertain demographics and interest in the material being presented. “Was the length: Too long? Too short? Just right?” some of the questions ask, or “Did you find this funny?” or “The amount of information was: Just right | Not enough | Too much.”

In addition to direct audience feedback, the Taster group built an analytics dashboard to offer the BBC teams testing their products more granular, realtime data such as view counts, engagement time, device and browser use, and location. The team is also looking at ways to improve the feedback loop for testers: “Right now you can take the survey, but it’d be really nice if we could then come back to you and say: Here’s what happens next,” Sharp said.

Groups testing out ideas can use all the data collected in the three-month trial to make a business case for whether or not the projects should move forward or get spiked. Three months has proven to be a long enough time to gather enough data to back those decisions, while keeping up a healthy rotation of new projects on the site for visitors, according to Woolard and Saunders.

“We would learn everything we could in three months,” Woolard said. “To some degree, it manages expectations, of the audience as well as the producer, that this wasn’t about spinning up complete new services that would have all the features right away. The reality is also occasionally you need a longer window to get access to different audiences or get to take advantage of different spikes. But broadly, it’s been a successful mechanic in managing expectations and understanding.”

Saunders and Woolard oversee modest budgets (“The cost of running the platform itself for a year is probably less than one of the premium episodes of one of our dramas,” Saunders said). A few ambitious projects around new digital formats that require work with external companies — VR experiences, for instance — typically cost around £10,000 to £50,000.

“Our proportionate spend at the BBC on content is huge in terms of television and radio — we make an awful lot of linear content. And at Taster, what we’re doing a lot of the time is giving teams who aren’t necessarily trained in data on a day-to-day basis clear data. Taster offers people a place to understand how to use the data they get,” Saunders said. “Most importantly, it gives people a chance to try different forms of storytelling out. Again, that’s hugely valuable in a world of hybrid content.”

Ideas can be retested, or tested in slightly different versions simultaneously: There are multiple projects around 360° video and virtual reality (live 360° video in Rio for the Olympics, a VR experience for Oculus Rift on the 1916 Easter Rising in Ireland). There are multiple interactive quiz projects (What’s Your Stand-up Style?, Find Your 17th-Century Love Match). When it launched, Taster ran multiple interactive video projects in order to toy with software like TouchCast and WireWax to determine whether or not to commit to storytelling formats using those technologies.

“I think the value comes from not piloting once, but piloting many times around an audience behavior, a technology, a format. That’s why at the bottom of the page you might see a bunch of projects all around 360°, because we can’t just do one — what will we learn if we just do one?” Saunders said.

“For example, we know we can reach a certain amount of people on television and a certain age of people. What can we do with hybrid formats? That is something Taster can help with,” he added. “What happens if we have a TV show about health and eating, that directs you online, where you can get your personalized diet guide there? What if we can create a really good quiz engine, pilot it eight or nine times with different shows. We might at the end be able to go to our TV commissioning colleagues and say: Look, we think we have an example of something that can get the TV show to reach more people online.”

Other outside organizations are interested in replicating the Taster model, Woolard told me, and “we’re working with university organizations, arts organizations, and others to open up the platform so they can also publish their experiments through us.” Taster has already been opened up, for instance, to the Royal Academy, for a 3D tour of its Ai Weiwei exhibition.

“They were using the BBC as a platform to engage with their own audience and gaining feedback, and we were gaining feedback on how people generally respond to 3D video,” Woolard said. “That is the direction we’re seeing increasingly: sharing out capabilities, working with like-minded audiences, and putting the audiences at the center, not just at the experience of consuming stuff, but at the earlier stages of feedback and development.”