The LeWeb conference isn’t very well-known in Germany, even it is one of the biggest and most innovative technology conferences in Europe. But one German startup founder knows the conference very well: Victor Henning, CEO and co-founder of Mendeley, a German science startup that pitched at LeWeb in 2009. Five years later Mendeley organized a pitch for science startups at LeWeb to establish a global network especially for them. At LeWeb talked with Victor Henning about the acquisition by Elsevier and the challenges of founding a science startup:
Tobias Schwarz: We are here at LeWeb because Elsevier and LeWeb have invited 15 science startups to present themselves at the startup competition Axon@LeWeb. But before we talk about that we have do to some groundwork, because science startups are something we don’t see very often at those kind of conferences. You are the founder and CEO of Mendeley which is a free reference manager and academic social network. What is Mendeley doing?
Victor Henning: The basic idea for Mendeley came to me as a Ph.D student seven years ago. I lived in Weimar, a small city in Germany, and I was troubling with all of the tasks that researches have, which is managing academic papers, finding stuff that I had downloaded, remembering what I had read, just keep in track of my research materials. I also did lots of research collaborations with other German and US-Universities, and we were always e-mailing our research papers back and forth and it was really complicated working with other people this way. So together with two friends I had the idea to start a company. The main question was: Why is there not a single software that automatically extracts all the relevant information from academic papers like the author, the title, the volume, the journal, the year of publication, the issue, just to help to you keep track with your own academic database? It could take all the PDFs that you have downloaded and collected, and automatically turns them into a structured sort of database. You could search in a full text database system, you could pull up any document. You could read it and highlight text digitally. You could use little digital sticky notes just for yourself as a memory device and keep track on your own research. And of course once it is digitally instructured, it would make it much easier to share with others as well. And that is what we did.
So in Mendeley you can start up groups, both private and invisible for the outside and public to share academic papers and see each others comments on the paper. This is very useful for both research and collaboration, but we have now also started to doing trials with Elsevier, the worlds largest publisher, which acquired Mendeley in 2013 about making the peer review process, which is the process of reviewing academic papers before they were published, making those kind of things more collaborative and more efficient by doing it in Mendeley, so people can see each others review comments on the paper, thereby Making the review more directive and afire quality. That was the initial idea for Mendeley;: helping people manage academic papers and help them share and collaborate around papers.
Another main idea that was really the one that got us starting the company and doing it on a big scale was the thought that if we actually have hundreds of thousands or millions of scientists using your software to extract information locally for themselves, why don’t we crowdsource and pool all of their information at the abstract to build an enormous academic databse on the web that is freely accessible to everyone? It could also come up with social information, because it is crowdsourced by the users. For example, if you upload a paper that you have read, we would know that your profession is journalist, you are interested in certain scientific topics, and so on. It is all anonmized, so you would never know the name or identity of the actual reader, but you would get some sort of general statistics. Using this information, we can track on a global basis the research trends in Biology, what Mathematicians are reading this week, or what is happening in Physics right now. And we have managed to build the worlds largest economic database in that way, which comes with the real time statistics people are working on and we have made that database available to others, not just on the web, but also through an API. That means that other developers can build programmes on top of that database under a Creative Commons license, so it is freely available to build applications on top.
That is what we are doing at Mendeley. We have 3.5 million users globally, literally on every continent on earth, so including the Antarctica Research Station. I am not sure if we have users on the ISS Space Station, but it would not surprise me. Our customers in terms of universites that pay for Mendeley subscription are some of the world’s leading universities like Stanford, Havard, or the MIT. We also have about 300 third-party developers, who have built applications on top of the free data that we provide.
Tobias Schwarz: That is a lot. When I started elementary studies in Germany, it had nothing to do with internet, there was no infrastructure. And then I went to an international university in Italy, I heard professors from Japan and the U.S. and it was a totally different way to teach and to learn. Did Mendeley start off well in Germany, or was it a success because you reacted to the global needs of the scientists?
Victor Henning: That is a good point, because all of us three founders are German. My co-founder Jan Reichelt headed his Ph.D. at Cologne, I did mine at Bauhaus University in Weimar, which is also where I met my third co-founder Paul Föckler, who is a computer scientist. When we started Mendeley, we deliberately wanted to start abroad. So we went to London to start the business there, because we knew we had to be good in english, which is the language of science and acedemics. Also London is probably the biggest hub for academic research in the world. There are a lot of universities around, you have access to a lot of venture capital and academic publishing has their headquarters there. But we felt that if we started in Germany and we got our own network of friends and academia into the network first, which are all Germans, people outside would look at it and think, this is a network especially for Germans.
I think it is very hard to convince a Stanford professor to join a German chord on chord network. And so deliberately when we launched, I did a tour on the US east coast. In ten days I presented Mendeley at Princeton, NYU, Yale-University, Havard, MIT, Dartmouth. I tried to get those people in first, and that is actually what happened. We built our userbase in the U.S. first and it is still our biggest userbase. About a third of our users are located in the U.S., then UK, which was our home base, and Germany followed later. I would say, unfortunately, because I am German. I think Germans tend to be a little bit slower in adopting these new technologies for some reason. In Germany there is a much stronger sense of privacy and the idea of crowdsourcing data and building a global database, maybe because of the privacy reason is more difficult, but in all other countries there was a very strong need that we served. That was the beginning, and Germany came afterwards. We also have paying customers in Germany now, we have collobrations with Max Planck Society, it is taking off as well.
Tobias Schwarz: Once you said in an interview that it is quite difficult to build up a science startup, because people put you in a niche. What were the problems and challenges you faced?
Victor Henning: When we started Mendeley and that was back in 2007 and 2008, I went and started pitching Mendeley to pretty much every venture capital firm in London. I probably did 30 or 40 pitches and I spent a couple of weeks. Usually I spent the 30 minutes of each meeting just explaining the system of academia. For example, that academics do not publish for money, but for recognition, how the peer-review process works, what the main points are, how academic publishing works in general. There was acually a lot of money in academia itself and academic publishing. But the reaction I got was ‘Yes, we see your point, but isn’t science nothing but a niche? How many scientists are there in the world? Maybe two to three millions? If you are lucky, you get 10 to 20 per cent market share. Maybe you have a market of 600.000 users that you serve and they dont have any money anyway…‘ – so that were the reactions I usually got, until we finally ended up taking investment from non traditional investers. We had the founding engineers of Skype who liked to invest, we had some of the founders of Last.fm, the music service who also liked to invest. Because there was a similiar idea which triggered Last.fm. Their business is about tracking people’s music lists and habits, and then build a global music database.
We did the same for research. Keep a track of people’s reading habits and then build a global research database. This is why we could win these investors, which genuinely sourced something with the idea and did not worry so much about the market being big enough. They supported us from the beginning, and by now we have outgrown these early predictions with more than 3.5 million users. Right now we are growing close to 180.000 users a month. We have students, we also have a lot of users in industry, in pharma, in biotech. We have a lot of users in government, even banks, law firms that do any sort of research tend to use our solutions. So it is a much bigger market than people initially thought. I think this is because we have successfully shown from 2009 to our exit in 2013, that you could build a solution for scientists that both can grow to global audience and make money by having customers like Stanford and Havard which pay 50.000 dollars per year for subscription, which also stand ultimately to a very good exit to Elsevier. This is an interesting market to build for.
Tobias Schwarz: Was that the time when it was all about the money? Did the investors see the potential to earn money with your startup? Did they have an academic background and understand why it would be useful to have something like that? I suggest, it is a very personal connection if you studied and experienced it by yourself and then someone came up with the idea that would have helped you with your studies.
Victor Henning: Yes, it was about the money. I think any good investor also has an alignment with you that it is both financially motivated but of course also share the same visions. And the investors that we had didn’t really have an academic background, except for one who was a Ph.D. All the others came mostly from the music industry. And I think the reason was that in music they saw something similiar has happened to what is happening to academic publishing, but ten years earlier. The trend leads towards digitalisation, the trend also leads towards crowdsourcing and what you can do with the power of crowdsourcing data. It leads to how to build a global music database, how to build up around that and also, how valuable that data becomes. And I think that gave them the idea that if you can replicate them in science, that would be really valuable. So they definitely did not just invest for charitable purposes. I think it helps if you have an idea that is bigger than just the latest e-commerce clone, which is very popular in Germany. You just don’t copy a U.S. company. If you have something that goes beyond that, and say our mission is to open up science and increase its efficiecy, and also to help the progress of science and humanity by helping researches work more efficiently, that is something that makes people feel good about, on top of making money.
Tobias Schwarz: When you see a startup conference like LeWeb: What is the main difference between a science startup and a technology startup? Are there any similarities that all science startups have in common?
Victor Henning: I would say, they are not inherently different from other tech startups. You encounter a problem that you want to solve. You are trying to get a prototype of the ground, you iterate, etc. What is different here is that of course the market tends to be smaller than maybe for some consumer or e-commerce startup. You are targeting a smaller group. But potentially you get much more revenue for users. So if you are a scientist and you want to outsource an experiment maybe you are willing to pay 1.000 or 5.000 or 10.000 Euros for that. You see how difficult it was for Whatsapp to convince someone to subscribe for one Euro a month. And that is the difference there, it is a smaller audience but there is a higher revenue per user. I would say, what they all have in common is that it is very difficult and a big challenge to actually find your userbase in the first place. And I would argue, it is potentially more difficult than for consumer startups, because the scientific market is quite fragmented. There is not really a lot of global media that covers tech startups for science or that can go to get your audience. With Mendeley I tried to attend both academic conferences and tech conferences, talk to traditional media, to the mainstream media as well as to tech blogs and to academic blogs. It was a sort of mixture that helped us gain an audience, but there is no one go to outlet, I think. So all science startups have to overcome that initial hurdle: getting the userbase.
Tobias Schwarz: When you talked about your tour on the east coast to the Ivy League universities, it reminded me a little bit of Facebook, which started its business at the same places. Do science startups try to build up the next BASF like many other startups try to be as successful as Facebook?
Victor Henning: I think science startups tend to orientate themselves towards the big successes in tech and consumer markets. You will see startups say: ‘I want to be the Facebook or LinkedIn for scientists…’
Tobias Schwarz: … or the Last.fm for scientists.
Victor Henning: …or the Last.fm for scientists, the Spotify for scientists, the Dropbox for scientists, the Evernote for scientists. You always have this model in mind to explain what you are doing. But that is for startups that build tools like Mendeley, they are building tools for scientists. There are of course other startups which are creating processes or doing primary research in biotechnology or else, and maybe for them its to be the next eMgine, the next BASF. Those ambitions definitely exist.
Tobias Schwarz: We talked about how Mendeley went international quite quickly, which is related to the needs of research and scientists. Did you notice any cultural differences in dealing with science startups in other countries?
Victor Henning: I think the reception towards these types of tools was definitely more open in the U.S. and also in Asia, especially Japan. When I went to Japan for the first time, and Korea as well, it had an amazing reception, it was strange. I was treated like a rockstar. I came to KAIST (Korean Advanced Institute of Science and Technology), which is sort of the Korean MIT. When I arrived on the Campus there was an enormous banner on the campus square with a photo of me on it and it said: “Presentation of the Mendeley CEO Victor Henning”. It was amazing. Then I came into a big auditorium and people afterwards wanted to take pictures with me. That was amazing, it was one of my favorite memories. But yes, things like that happen in the U.S. as well. From the very beginning, the Ivy League universities opened their doors for me, they organized sessions. In Germany, it was much more difficult to get that sort of response. I think in Germany it happened only once. We had the big userbase in the U.S. and the German universities saw that if MIT and Havard subscribed, this is maybe something worthwhile. In general I would say, Europe is a little bit slower to adopt than the U.S. and the UK.
Tobias Schwarz: In 2013 you were acquired by Elsevier which call themselves the ‘world-leading provider of scientific, technical and medical information products and services’. I can see why they were interested in an academic social network, but what are you doing now for Elsevier?
Victor Henning: We were in a position that was managed to retain our mission, our vision, our character as a sort of startup. So we were still on the same space we have been before we were acquired. We got a lot more resources. We were 50 people altogether, when we were acquired, now there are over 100 people working with us. We will be about 150 people in 2015, so we tripled our employees since the acquisition. And we are still focussing on the same things: making life easier for scientists, as in the academic workflow, in document management, in collaboration, in social networking. But because we are now a part of as you said, the worlds largest provider of academic information and science information, we are putting the two together.
For example: Elsevier has the largest citation database of which papers cite other papers and based on that you can calculate metrics. What are the popular papers, what are the authors, who have the most citations, and so on. And that type of information is very valuable for universities and scientists. For them, it can often determine the progression of their career; and those metrics will soon be integrated in Mendeley profiles for free. So if you are a Mendeley user you can keep track on your performance, based on your citations, which is for scientists the one currency they pay attention to. In return, we have integrated Mendeleys metrics,the real time statistics about the readership into Elseviers databases. If you go to Scopus now and ScienceDirect, which is the largest database for academic content, that have been verified by big publishers, you now can see the readership for this paper based on Mendeley metrics. That is the integration of data we already have.
But beyond that, as I mentioned earlier, we are experimenting around Elsevier’s co-business, which is publishing. And I should add: Elsevier publishes about roughly a third of the scientific papers. So particularly biomedical, but everything from physics to social sciences and everything between is published by Elsevier. It is a third of the world’s scientific content, every year. And we are now trying to make that process more collaborative, the peer-review process, by inviting the reviewers into Mendeley-groups, where they can see the other reviewers’ comments. This has achieved very good feedback. The reviewers say, they find it very useful to see the other reviewers’ comments. Sometimes there can be resolved an upcoming conflict or an open question before you pass it back to the author. That saves us time and proves the quality you give back to the author. It also creates a much better experience to everyone involved. We are also collaborating around new ideas on open access, so using the Mendeley platform for open access publishing distribution. So hopefully we can share more about that in the future.
Tobias Schwarz: Elsevier and LeWeb organized a presentation just for science startups. What have you been expecting at the event?
Victor Henning: The one thing that I expected at this event was to connect the science startups with each others. When I built Mendeley, the one thing I always wanted was to meet all the other founders in the science and research space, even if they were competitors. I also tried to build relationships, so that we could exchange experiences fromthings like where the market was going to what the major trends were, what problems they faced, how they overcame these problems. We just wanted to exchange our experiences with fellow founders. Since the acquisition by Elsevier last year, I got a lot of requests from startup founders to become a mentor for them. They have seen me build up a company from scratch to a bigger position. I thought the best way to actually do this is to form a network of science startups, where the founders can talk to each others and I can give mentoring and advice to all of them. I wanted to connect them with each others, connect them with Elsevier as well. That was the goal of yesterday’s event.
The new network that we announced yesterday is called Axon. It is thought to be the network for science startups, which is to have the founders of the most promising science startups, being able to connect and exchange information and experiences. With the competition we invited science startups to apply, to present on the event. I picked them not so much on how big they were or how much funding they raised, but how much I think they had the potential to change science and scientific processes. Some of them are in a very early stage, they have just launched their beta, but I think they all persued big ideas. If they come to fruition, they could make science more efficient, more open, more collaborative; that was the main criterion for me. And so I hoped first of all, that we can get these science startups to talk to each other, form a network, and second of course we can form an audience.
I reached out to LeWeb and said, I want to do an event with really interesting science startups all over the world. Could we do something where we get them all into a room and let them present. That finally happend yesterday on the conference. I think we had a really good response from the audience. We received like a lot of emails afterwards which said things like ‘It’s great to see such cool innovation in that space.’ I heard from the startups that they really enjoyed it and there were some journalists, too. I think it was a really good event.
Tobias Schwarz: Some weeks ago I interviewed a CEO from a Silicon Valley startup and he asked if he should go to LeWeb. What were your reasons?
Victor Henning: I had several reasons. Mendeley itself was part of the LeWeb pitching competiton back in 2009 and I have really good memories of the event. We were pitching to a big audience, there were a lot of American investors in the room that we formed relationships with afterwards. Plus: I love Paris and the event was a really good event. And I enjoyed coming here. And so I thought, lets try to replicate that and bring the other startups here. Besides, LeWeb is part of Reed Midem, which is a sister company of Elsevier. So I thought we could work together and I got a lot of support from them, blogging about the event and getting some free tickets to the startups event.
Tobias Schwarz: 15 startups were here yesterday, but how many startups applied for Axon@LeWeb?
Victor Henning: There were about 30 startups.
Tobias Schwarz: Eight startups were from the EU, six from America and one was from New Zealand. Does this reflect the status quo in the science startups scene?
Victor Henning: I think it is fairly representative. There are probably more science startups in the U.S. than in the EU, but this could also has happened because we did the event in Europe and it was put together on fairly short notice. We only annouced the competition about six weeks before the event and then the application phase ran for two weeks, so we are very happy with the response that we got. According to many of the U.S. companies, I think it was too short notice to apply and to come to Europe. Next year I hope we are able to annoubce the event in advance and get more of the U.S.-startups to come over. But overall I think it was fairly representative.
Tobias Schwarz: Lenny Teytelman from ZappyLab wrote about yesterday: “I support enthusiastically Victor Henning’s effort to showcase science startups. In a sea of interesting general-audience companies at LeWeb, there is no way that an amazing company like Publons (opening up peer reviews for all to see), can get featured at the main event. Therefore, it is important to set up the precedent for an event that highlights science startups to the right audience of investors, reporters, and others in the science space. I am glad Victor did it; I hope it’s not a one-time event; and I want to help it succeed as much as possible.” It seems that you did a good job. Last, but not least, the European Commission talks a lot about startups and is working on Europe’s Digital Agenda. What should the politicians be aware of regarding science startups?
I think there is a lot to be done on helping scientists launch companies. When I was a scientist and my cofounders were also doing their Ph.Ds and we were going to get if off the ground, there wasn’t really a lot of support from universities and the feedback that we got was almost discouraging. So we had the technology transfer office of the university of Cologne looking on our idea and the were saying: ‘Oh, it is already out there. We cannot help you with any of the IP and you have to be very careful, because there are already tools that can extract data from paper or from books. There are visualisation tools, there are document management tools, so what you are doing is not new.’ And I think this was not very encouraging feedback.
As I said, it is still not very easy for science startups to raise funding, because there is this perception in the VC community that it is not a big market. If the EU wants to directify that, I think there are two ways to do that. One is to provide more guidance to scientists of how they can start a company. Probably the best way to do this is to try to get more entrepreneurs to University and also to talk to the scientists. They should tell them about the important steps to start a company. They could also get together with non-scientists, maybe business people or experienced entrepreneurs who want to start a new company, which is not e-commerce or a clone, but something generally exciting based on scientific breakthrough. Put these people together and help them form teams.
The other important thing is the funding situation. In Euope it is generally more difficult to raise funding than in the U.S. The european investors are more cautious. They raise smaller amounts of money compared to the U.S. companies and they adress that by public-private partnerships. Compared to the the UK, for example, the government gives money to VCs to invest on their behalf, and we had one of those VCs invested in Mendeley. That would be a good way to provide more funding to these startups.
Tobias Schwarz: Thank you for the interview.