Science Exchange Elizabeth Iorns
So, um. We started thinking yesterday about what to talk here. I thought about the different aspects of bringing the tech ideas into the science space. Some of the key points that Joseph wrote about was catalyzing the formation of teams and companies that positively transform science. The emergence of a new type of science entrepreneurs to .. effect contemporary relationship. That's inspirational and inspiring and it's exciting to think about how to encourage scientists to consider this alongside the academic careers that we all know about.
I came from a traditional academic career path. I did my PhD in London and I did a postdoc at University of Miami. I was ddoing academic research until I realized that there were problems that I could address with technology, and it was more exciting than running a lab. So then I founded scienceexchange.com. Today I hope to share some lessons about starting a company that might help other scientists with the transfer. I am really happy to help people make a transition to this career.
Just a brief introduction to what Science Exchange is.. it's an online market space where people pay for services from institutions. It wsa for building access to expertise and infrastructure in a marketplace. Instead of a local core facility at your own institution, what about a portal where researchers could transact on expertise available outside their institution and globally.
So we were founded last year in May, we launched in August, we are based in Palo Alto and took part in the ycombinator S11 program. We built our product quickly, we raised a seed round of financing including Jason Harowitz, and Cross Sight amongst others. So we now are used by more than 1000 core facilities across the United States. There are some nice brand names like MIT, Harvard and Stanford.
We have critical mass and we have users accessing core facilities they wouldn't have had access to anyway. I don't want to talk about this. I want to talk about why I created Science Exchange. I would be happy to answer questions about creating a startup in this space.
You need to solve a real problem. Often there is technology searching for problems, especially the young guys coming into ycombinator. I was one of only two women out of 164 people in the program. Almost everyone was young, white males, all engineers straight out of college. They are so talented and brilliant- some of the smartest people I've ever interacted with- they were looking for problems and weren't sure what to use these skills to solve. When thinking about building a startup, you need to research the market and maek sure that the problem exists. If you are going to make a tech startup, building a small one is just as hard as a big one so you might as well solve a hard problem.
So the Science Exchange mission was that core facilities is increasingly important, but access is just as hard. So we talked with a lot of scientists, and I was able to understand as a scientist that it was a problem that people experienced on a daily basis. So we were able to get numbers on $100B spent on research every year, much of that through things other than core facilities, but increasingly so.
So, do something you're really passionate about. I love this quote from Paul Graham. The best problems to solve are the ones that effect you personally. You have to understad the problem that you are trying to address. Often this can be hard when you are starting a company. The problem you are solving has to be important. You have to have that connection. For me and my own research, my credit card balance was effected by the lack of a good marketplace.
Make a start. It's hard to make that first step, that initial step into this world, especially if you're not from the tech community. It can be really scary, and you might not know what to do to start a company. But one of the things that... and get on with building a company.
You need to share your idea. One of the problems was that people thought their ideas would be stolen. You shouldn't tell anyone, and my cofounders were both in the tech startup world. No, that's not the way you do it, tell everyone about your idea so that you can get feedback and build something that people really want. That was a valuable lesson. We started with market research and we thought about the problem, and whether people found this to be a problem, and then we applied to ycombinator. That was from February 2011 to taking ycombinator in May. Making a start and making progress was what allowed us to make a successful company.
Surround yourself with the right people. If you actually look at all these companies, these startups were often done by a single founder. When you're looking for investment, or skills required to actually get up and running, it's important to have diversity in that team so that you can handle any of the problems that you will be addressing. Science Exchange has 3 very different founders. I am an academic scientist, we have a technical engineer and a business person. That diversity has really helped us with understanding all sorts of different problems that has helped us with building a company.
Do things that don't scale. If you have a really complicated accounting, dealing with transactions, I have to build this technology around that.. and then being a part of ycombinator is valuable because Paul was like, that's a great problem to have. You should just get started, start taking money manually, and then you will figure it out when you need to. Only scale when you need to. Doing things manually was actually really important, to understand our customers and see what they needed, and that when we needed to scale that we were doing so in an important way. Customer service is super valuable when you are starting.
The sixth point is to listen to your users. This goes back to the customer support. This is a really interesting quote. Customers are used to companies ignoring them. Some startups have fanatic loyalty from their initial users.. might not have the best product, or might have a shitty website with errors, but they reach out and engage with the users, and that personal interaction creates a community around the product that is extremely loyal. That was something really valuable from the start, it was making sure that when they tweeted that we would reply to them, we were calling them and visiting them and making sure they feel involved in the community of Science Exchange.
Don't always do what they say. If they have a problem, you should listen, but their solution might not be the best solution. Your company is your company, and it is up to you that to figure out how to solve the problem most optimally.
The seventh one is iterate quickly. This is true of startups in general. Iterating enough times until you find something that works until you run out of money. That's really a startup. I think when you have this really early users that interacted with you and are your community and driving your growth, you can listen to them, they want you to speak and improve quickly, they don't want you to just sit around. So, that's really important for iterating quickly.
When we first launched, our product was post a project and we would find someone to do the experiment for you. That was a frustrating experience. But we physically went to the person, and we checked if the person was able to do it. You don't know if anyone would be able to help you, we switched from this auction model to an order model. We built a database where we exposed the value from the start, we had the ability to see and filter, see the feedback on using those core facilities from other users. So exposing the site right from the interaction of the site, and that was frustrating for people who were posting projects, but it helped providers because they didn't have to post and miss out on work. So when someone was actually requesting, then they were more likely to actually order and they already went through that process.
Get the right investors and advisors. This is something that I never realized before coming to SV. When we talk about the type of startups that we're talking about today, they are around infrastructure changes and technology, they are not the equivalent of biotech startups. We're facing a marketplace that is similar to Airbnb or Odesk that are very different from the problems that Genentech faced. These are very different advisors from Biotech companies. THis is probably true of other scientist startups. ResearchGate has more similarities to Kaggle or Faceboeok, not the other sites. PeerJ is more similar towards sites like youtube or spotify that have to license content from providers.
For us when we were thinking about adivsors, we looked at talents that could help us with an online marketplace and understand the things we need to put in place. We had Jason Horowotizz, including the OpenTable person, and the Airbnb person and the former CEO of Ebay, and this helped us figure out our problems.
The ninth lesson is to commit. I think this is something that is very hard to do. It is very scary that you need to have these backup plans, I need to keep my day job and all of tehse kind of things that you put in place where, if it does fail- which it will, by the way- that you are not kept high and dry. I think that's a self-defeating attitude. If you truly believe that your startup is a good idea, and you're executing on it well, then you need to commit to it and go to it. There is a limited runway and enthusiastic energy. When you start, you get enthusiastic, make the phone calls and do all the hard things that are not fun. Over time you will lose that energy, so you do need to make progress so that you don't lose your energy, otherwise your startup will fail.
Finally, you need to have fun. At Science Exchange, it's the only time you will ever have to work with people you choose with and to do the things you want to do. We want to make sure our team is happy with each other, we're good friends, we do things outside of work as well as working on the startup. That's what we focus on at Science Exchange as well.
I am really happy to make introductions for people, or giving advice and doing science startups. You can reach me at elizabeth@scienceexchange.com.
It was interesting that you compared yourself more to Airbnb or Peerj or Ebay or something like that. That's really interesting. Could you explain on what you see, what are the core similarities between Science Exchange and Airbnb?
For us, Airbnb is a company that we looked to for inspiration. When you think about airbnb, we think about people renting out their house to strangers. That's a pretty scary prospect. They have to make sure that people trust them, they have to build a community not afraid to stay with strangers, and it's the same thing with Science Exchange. It's not about selling something online, it's your job- it costs a lot of money- these transcations are thousands of dollars, so that ability to have trust in the site, and that we look at verification and connections between our researchers in the facilities that they work at and all of thsoe aspects that we look to Airbnb for andd their problems, and what can we do at Science Exchange.
To tie that back to the theme of today, what do you think the effects are on the overall marketplace of skills would be, if there are more people taking that up? Could it make more efficient science? What is the opportunity there?
It would make it a lot more efficient and a lot faster. There is this culture in science to do everything yourself, or to learn each experiment,a nd a lot of the time that will be related to quality issues in academic research and reproducibility. Using core facilities can help on these core points. These people do these experiments all the time. That's all they do. They don't have to publish, they just provide data back. They don't have an interest in what the outcome is. There are none of the original perverse incentive. Instead of outdated revenue model, it's a parallel approach to funding science.
Quick question: when you were raising money, did you have to envangelizing science and explain the flows of moneys, did you have to educate the investment community about science? We find that they don't know science.
I think that VCs don't know about academic science. For me, I was really lucky that there were people like academia.edu that had gone out and rallied the VCs about the academic world. We weren't the first ones. For us, it's all about the economics and doing a value proposition for them. It's very easy to sort of go in there and say this is how much money is in academic research, these are the trends, this is where it's being spent, here's the collaboration for core facilities, so going out there and .. from an understanding of the academic research world, just say look at it as a market and an ability to disrupt a market. That's how we convinced them.
Could you share what your model is? It's our payment platform. We are totally open access, we're free for core facilities to list, and free for researchers. Our whole thing is that it's too hard for people to transact money between institutions. The big barrier was that when you work with a university, you can't just pay money to anyone, you have to have an approved vendor to receive that money. That process is very time consuming, and there's these university purchasing system. So we ewent throug hthe university purcahsing system. So the researchers can now use the researcher somewhere else, and figure out how they are going to transact otherwise. So then we charge a transaction fee.
From the standpoint of justifying expenditures on a research budget? No, that's an issue. We've never had an issue in terms of service in terms of how much to allocate to different things. They have a lot of discretion in how they spend that money- they could buy the reagents and do it with their own research staff, or they can use their core facility or some other core facility. Yes, we're totally open, we want to be a central market, so we're looking for anybody to find anything. It's not restricted at all. We verify everybody who is a provider, but you can be a private consultant, you can be a CRO, you can be a commercial service provider or an academic core faciltiy. Yes, all members are vetted. We have a manual approval process for providers, we physically speak with people and check out every facility on our site, so we create the content on the site. We want to make sure that our experimental classification system si correct, and that's really hard if you have community generated information, so it's really hard for them to know what they're buying, so we just do that manually.
Thank you for a very nice talk. You are solving a lot of dilemmas for the research community is that, I hope this is part of your model, you're very talented and bright, the ability to notice how much effective flux is going through different parts of research is a wonderful metric for being able to talk about the monetization of the science budget... if scientists forget, but financial anylsts like myself, the velocity of money, the fact that you put so much into a particular experiment or somebody performing one of your services, that creates an opportunity for someone else to do the same, so it speeds up, so it advances it internally, so you can do a velocity indicator for turning things around. THis is very attractive to the funding cycle as well. It allows people to fund more quickly. So if I add a little more, it goes down the same path and now it's a big industry.. so communicating that back and publicizing that as an effect, may cause a tremendous attraction from the financial inddustry... historically, that's how industries surveyed each other. Scientists don't tend to exploit that.
We try to track, at a very detailed level, around the use of our site. Metrics. In terms of value proposition that we offer is the first time, where money is being spent on experiments. Not just local, but globally. It's valuable on that level. When you look at the published literature, only 1% of experiments are published. Which money is being spent where? This is what funding agencies are interested in looking in. They know that most people are spending money on which experiment now, and then target them directly in real time.