Sciencebite Is Metasearch For Scientific Expertise

A great startup idea often has a chunk of science at its core. But getting the discrete worlds of industry and academe to combine is not as easy as it could be. So says Berlin-based Sciencebite, a startup co-founded by Oxford- and Cambridge-educated scientist Samson Rogers who spent 10 years working in academia and consulting for industry before deciding to make a business out of strengthening those very same links.

“We want to take applied science out of the paywalls and put it in touch with industry,” said Rogers, presenting on stage during the startup battlefield competition here at TechCrunch Disrupt SF.

[gallery ids="1053609,1053610,1053611,1053612,1053613,1053614,1053615,1053616,1053617,1053618"]

Sciencebite’s plan to oil the wheels of industry-academe collaboration is a real-time, algorithmically powered matching platform that makes it quicker and easier for industry to find the right scientific expertise to get a project off the ground. It’s not going after lengthy consulting engagements involving commercial players outsourcing an entire project to a scientific team. Rather, its focus is to become a platform where experts can be found to provide personal, one-on-one advice on specific questions to help kick off a commercial project or spec out an idea. Think of it as a highly specific Quora for scientific expertise.

It also wants to become a platform where scientists can raise their public profiles generally, so that it’s easier for industry to find them. Sciencebite is today launching its matching platform in open beta.

Of course it’s entirely possible to track down an expert to advise on an R&D project yourself, armed with digital tools like LinkedIn and Google and a bit of time online. But Sciencebite’s metasearch cuts out some of the legwork by narrowing the search to experts in a niche area from the get go.

“We let scientists, for instance R&D professionals, enter key terms or search terms into the Sciencebite search engine and we crawl and we analyze academic literature online. We apply our algorithm on it and we are able to identify the best experts, the best scientist, who can address that technical query,” explains co-founder Stéphane-Bertin Hoffmann, discussing the Sciencebite search process.

The company is starting by conducting weighted keyword matching of expertise queries with academic data published online but obviously hopes to build up its own knowledge base by attracting scientists to use its platform and flesh out their own profiles on Sciencebite.

It doesn’t need mass sign-up of scientists to begin functioning, though, with Hoffmann noting that it’s able to obtain contact email addresses for experts via the abstracts and papers they have published online. That means there’s no need to gather a critical mass of experts before the platform has the chance to become useful.

Another advance over a manual expertise hunt (via LinkedIn et al.) are the additional features built in to the Sciencebite platform that likewise aim to take the strain out of knowledge cross-pollination.

[gallery ids="1051423,1051431,1051426,1051430,1051429,1051427"]

For instance, the platform gives scientists the option to respond privately to a request for their expertise, and at that point it also becomes a marketplace where they can charge a fee for carrying out the requested work. (In the future Sciencebite hopes to establish set market rates for these fees, based on factors such as industry type and engagement, but initially scientists will be able to set their own rates.)

Or they can respond publicly to the request, sharing their knowledge with the Sciencebite community but forfeiting the chance of a fee if they do so. They might choose to share something publicly to raise their profile on the platform which scores contributors based on participation and reviews — so they have more chance of being foregrounded in future searches.

If a scientist can’t or doesn’t want to answer an expertise request query themselves, the platform also lets them refer it to a colleague who may be better placed to help — and earn a slice of any payment for resulting consulting work. Sciencebite’s business model is to take a 10 percent cut of any transactions agreed via its platform.

“The user can click on the scientist we provide, can review the scientific literature backing the profile, and he can get in touch with the scientist directly and send a question. The difference with just using a search engine, or maybe using Google scholars is that we are actively connecting both sides and we are incentivizing the scientist to support and help with the question,” says Hoffmann.

“We are targeting specific know-how, niche know-how owned by an individual scientist. And this know-how we expect it would usually turn into one, two, three days engagement, which leads us to fees in the area of €2,000, up to €4,000,” he adds.

The platform is tackling IP issues that might prevent collaboration by proving a standard NDA agreement that scientists and those requesting their expertise can quickly and easily utilize to kick off the knowledge sharing. Sciencebite also allows for proprietary NDA agreements to be used in the event that the standard NDA is not sufficient.

In terms of competitive landscape, Sciencebite competes with established expertise matchmaking businesses such as GLG which take a human-powered approach to the problem. So it’s claiming advantages of speed, scalability and impartiality vs those rivals’ more manual approach.

It also has cross-over with open innovation platforms, such as Innocentive, Ninesigma and Kaggle, according to Hoffmann, but again hopes to stand apart because of its real-time matching element.

The real-time matching element will not necessarily lead to instant results of course, given that the length of time it takes a scientist to respond to a request for their expertise is not something Sciencebite can control.

[gallery ids="1056403,1056402,1056401,1056400,1056399"]

Academic tardiness in responding to email is certainly a risk for its business. But it’s hoping to counteract that by incentivizing responding on the platform with social ratings, such as scores and reviews, that amplify the visibility of those who do engage in a timely fashion. And of course by building a low-friction route to market for scientists to sell their expertise.

“[Response times are] a challenge we’ll have to deal with, at the beginning especially — when the platform is not seen as a reference for those sort of issues,” concedes Hoffmann. “But we are really offering something of value to the scientist who has been contacted. For instance the user who send the request has their identity checked [by Sciencebite] because they have to sign in with LinkedIn. So we make sure that those are real requests, from real persons in industry. And we provide them with a specific question, specific need — so there is a trust created with a connection.

“And we also capitalize on the fact that scientists, especially in academia… they really do want and seek connection with industry. They are looking to promote themselves, promote their expertise and they don’t have, we believe, the right tools to do so so far.”

Sciencebite has been funded by Berlin-based angel investors thus far, and is now seeking additional seed funding to grow its team and make good on the vision. Areas of focus for new funding will be beefing up the technical team so it can improve its matching algorithm and partnering with R&D companies to get the tool in the hands of people who are going to want to use it, says Hoffmann.

Judges Q&A

Q: There are a couple of companies that are connecting high end whitepapers with industry and individuals. Do you measure yourselves against those companies?
A: The relevant competitors are The Open Innovation crowdsourcing platform… Nobody has yet got the right model for bridging the two sides. We’ve got the right model… Most companies simply don’t use the Internet for this… they hire offline brokers. The whole communications take place offline

Q: You had a very simple question you were asking [in the on stage demo]. In the real world most of the questions that stymie a team aren’t so simple. Have you tested this with R&D teams?
A: Yes we have. We’ve been in private beta for the last few weeks… that question was related to a real problem that I had. We’ve had people with immediate needs asking questions and scientists have replied

Q: It seems like if you just focus on the heavy R&D science and scientists, rather than this piece of it which feels like a consumer business I feel like those are two pretty different markets. The connecting science bit feels like it would be less of a Q&A based function, less about profiles and more exclusivity to it. How have you navigated that or is it too early to tell?
A: There’s a bit of both. There’s a profile which has my credentials but actually more important in commercial R&D is establishing this very specialized answer to a question… They need an answer to a question and there isn’t a way to do this.
If you compare this to software development where there are well developed tools online and communities that share answers… There’ s a joy to share knowledge in software development and it’s a terrible experience in science R&D

Q: Again that’s very transactional. Developers have had years to build reputations in profiles… Where you guys are willing to create motivation for people to answer questions to create scale seems like an area where you have work to do…
A: Indeed. That’s why we’ve focused on this matching tech to find people to answer specialized questions

Q: What happens if a simple question requires a much longer answer and the expertise turns into a consulting engagement? Are you involved in that process?
A: Consulting engagements are what we want – especially short engagements for know-how. A few days work, for which we take 10% [of the fee]. That’s our main business model

Q: Aren’t there IP issues involved with sharing?
A: IP and confidentiality are very important for our customers. We design that with those in mind. In the team we have a background in R&D… knowledge brokering, consulting and R&D funding… The way we’ve set it up for the clients is with this in mind

Q: What’s a typical fee?
A: Two to five thousand dollars

Q: People would pay that for the answer to that specific question?
A: Typically one would expect to establish some relationship to discuss the matter but almost any question has a lot of considerations and that’s something that would involve a few days work

Q: What is the frequencies for companies to solicit external expertise?
A: That varies a lot. Some 5% to 10% of R&D budgets of large companies are outsourced for know-how based services