Perhaps it’s no surprise for the Harvard-trained virologist, who traded in a promising medical research career to launch the social network for scientists. But still, in a world where the impact of social networks is usually measured by how many news headlines they can generate, he prefers success stories that have a more direct impact.

Take the example of Rafael Luque, a chemistry professor at the University of Cordoba in Spain. Luque found a collaborator on ResearchGate that he’d have never come across in the real world: a post-graduate student in the Philippines called Rick Arneil Aracon. After using the site to connect and discuss some ideas, together they discovered a novel new method of helping to make biofuels from the leftovers of corn cobs.

The technology is still in development, but it’s evidence of real impact for the site — and a hint at the substantial change that’s happening in the way scientists can work online.

“When we started, people told me you have to get all of the big professors on board,” says Madisch, as we sit in the company’s Berlin headquarters.

His answer was precisely the opposite.

“No,” he says. “If you get all the people who will be professors then it will succeed. We have people who joined four years ago who now say they use ResearchGate for communication in their lab: that’s what I want.”

It’s been a big few weeks for the network, which recently announced that it had broken the two million user barrier. That marks a serious milestone, even if it seems small in comparison to the billion citizens of the United States of Facebook.

It may be dwarfed by Zuckerberg’s empire, but Madisch and his team — including former Facebooker Matt Cohler, who sits on the company’s board — think that they can punch way above their weight with a much smaller community. Why? Because their couple million users are all professional scientists and academics who are all trying to change the world.

So what’s next?

“The next big thing,” says Madisch, “is reputation.”

“We want to try out a community review system: I’ve had it in mind for years now, and I want to see if people will accept this way of reviewing data and sharing data.”

In the end, he wants ResearchGate to be a place where researchers publish their papers, make their full results accessible and gather reputation. That in turn, he thinks, could make the site’s ratings a reference for outside funders, governments and the science community at large.

“My goal is that at one point I’ll talk to the larger funding agencies, and if they are accepting our score at some point, every scientist can think ‘should I publish in Nature and pay a lot of money for it and only get reputation for part of the data I created’ or ‘should I publish it all on ResearchGate and get reputation for everything?’.”

Most German startups are ‘shit’

It’s a big ambition, and one that goes way beyond the scope of many German internet startups — many of whom are happy to coin it in with clones or copycat services.

This is a problem, says Madisch, and he doesn’t mince his words: whether or not unoriginality is ethical, he believes the fact that so German startups are built by MBAs rather than passionate experts is dragging the country’s web scene down.

“Most of it is shit,” he says. “If you look at the people who are founding companies in Germany, you have many people who come from the business world, and they think as business guys: What models exist? What can I do efficiently the same way without changing a lot? If you look at other countries, especially the U.S. or England, the people are coming from the industries: they know what they’re doing.”

He feels particular kinship with the other exciting consumer-focused startups graduating out of Berlin’s over-hip scene: the likes of Soundcloud and Gidsy. They are building services that do new things, not copying the examples set elsewhere — and they have a lot to share with each other.

“We talk to each other,” he says of the cadre of CEOs now operating in the city. “We talk and we meet regularly for coffee or dinner, we’re somehow in constant conversation if we need something or if we think we should talk at some point.”

Still, ResearchGate has benefited from the clone culture, albeit indirectly: it hired a significant number of staff members from StudiVZ, the Facebook copycat that imploded spectacularly earlier this year. Their loss is his gain, and it’s perhaps a sign of what might come after a generation of engineers graduate from building copies and start branching out on their own.

But while a big vision may be unusual for Berlin, ResearchGate is not alone in its ambitions to upturn the stuffy and defensive world of scientific publishing.

Mendeley, another scientific startup with German origins, is also trying to use data to disrupt the existing academic structures.

However, the fact that there are two major rivals in Europe both around the same size is not a concern to Madisch.

“I’m more afraid of the ones coming than the ones who exist already,” he says. “And I’m more concerned about what the publishers will do in the future than really having competitors in the online world.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Flash analysis: future opportunities for Pinterest
• Monetizing music in the post-scarcity age
• Facebook’s tactical retreat on privacy

Co-founder and CEO Victor Henning told me that the company’s API, which offers other services access to its trove of millions of scientific documents, has just surpassed 100 million calls each month. A year after launching the service — which provides access to the information stored in around 65 million scientific papers, documents and files in Mendeley’s databases — the site has around 240 apps that employ it, and is now seeing growth rocket. And that growth, he said, comes through the increasing popularity of those apps, not through any new, specific effort on its own part.

Henning highlighted a few of the projects which are creating the bulk of that traffic, including some interesting examples of the benefit opening up some data can have on what is, traditionally, quite a closed market.

For example, Readermeter.org and Total Impact are both services that measure how much impact a particular scientific paper or author has by analyzing how much and how widely the work is read. Meanwhile productivity app Hojoki integrates with Mendeley, pulling updates in alongside other services to create a personalized newsfeed for you.

Perhaps most interesting of all is OpenSNP, a project that allows people to share their genomic data with each other. It’s using Mendeley to help users cross reference the data they are finding out about their own genetic makeup with the latest scientific research as a way of understanding what’s going on in their bodies. If that’s not mind-blowing, I don’t know what is.

Henning says this is all part of a move to opening up science.

Next up? Wary of Twitter’s recent API troubles, Henning says that Mendeley wants to enable as many third party apps as possible, and has no plans to force money out of them — instead focusing on paid accounts and secondary services like its institutional dashboard.

“In academia, everything revolves around journals — but everything is behind a paywall, only available to universities who pay expensive subscriptions, and without APIs or other ways to access the data,” he said. “People see the importance of this [Mendeley's API] for opening up science.”

“We want to keep growing the ecosystem, get apps talking to each other and get more integration with the Mendeley experience” he adds. “But we don’t intend to monetize the apps.”

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Facebook’s tactical retreat on privacy
• The 2013 task management tools market
• How consumer media will change in 2013