#FoundersFriday with James Harwood from Penelope Research
We are running a new recurring series on our blog, #FoundersFriday, in which we interview the founders of different scholarly communication businesses, asking them to share their advice for others and their perspective on the industry as a whole.
What made you decide to leave academia and launch your own idea?
I think, in a way being a scientist is very similar to trying to start your own business. You’re driven by a similar desire to do something new, to try to figure out problems all the time. The two overlap pretty well so it never really seemed like such a big jump.
I was working in research, I was doing neuroscience research in a hospital in Germany, and I started getting frustrated with seeing work that was published that wasn’t as good as it should have been. Research where the statistics weren’t reported properly or I couldn’t get hold of the underlying data. That always frustrated me but I think working in a medical setting, research was no longer just an abstract thing, it’s very real. All the research that’s done badly is research that’s not helping these patients. So then those frustrations start making you feel pretty mad, and when things start to annoy me I’m not very good at letting them go. That’s when I started to think about how machine reading could be used to improve what’s being published and started talking to publishers to try and find possible solutions.
If you could go back in time and give your pre-startup self one piece of advice, what would it be and why?
I think the biggest thing that I struggled with is the tendency to want to be a perfectionist and to keep things private until you feel like they’re as good as they can be and only then release it. Which I guess is something you do in academia, you tend not to release half-baked ideas, but in a startup, you really just want to iterate quickly, get ideas out, get feedback, make them better, otherwise you waste a lot of time. So my advice would be, don’t be too much of a perfectionist!
Suppose I have an idea for a tool, or a solution for a problem, within the research landscape and I want to develop my idea into a business. What would your advice to me be?
One thing that I’ve seen amongst people that want to start a business is that they run away with development really early on, they try to get a technical person, they start making code and they build something, without really testing the idea in the first place. So my advice would be to not build anything. If you can visualise it, just make mock-ups, make something that looks like it’s real and then go and present that to people and get feedback. You’ll always find that it’s not quite right, you’ve got to change it, so it’s better to have done that before you start building any code.
In the scholarly communication & publishing space, it’s often said that people have the same conversations again and again. Can you think of an issue that, in your view, people aren’t talking about enough?
Now that I’m working in this space and I attend conferences to do with research integrity, I find that those conferences are really well attended by academics and by publishers, but not so much by funders and that surprised me. Funders don’t seem to be as engaged as the other stakeholders, even though it’s their money that’s being spent on all this research. I think also there’s an issue around not really knowing who’s responsibility it is to make things better. Damian Pattinson, at Research Square, he talks about the “triangle of irresponsibility”, where academics say that it’s the publishers that have to change, publishers say it’s the funders that have to change, and funders say they’re not going to do anything until the academics want to change.
What are you most proud of?
There’s this group called the EQUATOR Network who advise publishers on the golden standards for research reporting basically. I’ve been doing a lot of work with their UK centre in Oxford. I’ve always been a massive admirer of them, because I think what they do is really important and recently they made me a fellow, which I was pretty chuffed about.
You were awarded a Digital Science Catalyst Grant. How did you hear about the Catalyst Grant?
I had already found out about Figshare, Altmetric and Overleaf, so then I found Digital Science and then the Catalyst Grant. But it was also brought to my attention through the UK Government Entrepreneur Network I’m in.
How has the Catalyst Grant and your relationship with Digital Science helped you?
The Catalyst Grant is quite unique as a source of money. There were two traditional sources of money that I could have tapped and neither of them were a good fit for me. There’s the typical academic grants, but very few funders fund research on research, or the development of new services to improve research. They don’t really want to fund that, they just want to fund the research itself.Then there’s the typical business investment, angel investment etc, I think a lot of them think that academic publishing is too niche, they don’t understand it, they definitely don’t understand the weird business models that exist in the academic publishing world, so it’s not an easy sell to try and get investment.
Digital Science money is kind of unique in that Digital Science understands the industry really well, its problems and its needs. They’ve made a pot of money to get new ideas off the ground. The fact that it’s ‘no strings attached’ means that I had a lot of freedom to grow in ways that I didn’t necessarily know I was going to.I think that’s really important at the beginning, if the funding comes with too many strings then you’re shoehorned into only exploring one idea and if it’s not the right route then it makes it quite difficult.
What does the future have in store for Penelope?
In the immediate future we’re going into private beta, we’re piloting it with a few publishers now, that’s great. We’re piloting a tool for authors but we’ll also be testing a tool for editors. That’s the immediate future. In the mid-term future Penelope will get smarter and do more complicated checks than the stuff it’s doing at the moment.