Helping the unaffiliated researcher

How did DeepDyve come about?

Our founders were originally bio-scientists. Their careers began in academic institutions where they had very good access. When they left academia and went to work for large companies, access still wasn’t too much of a problem.

The problem really came when they left these big companies and joined start-ups. Then they found that journal subscriptions were very expensive and what they used to do – download a batch of papers and browse them – was no longer possible. Buying 100 articles to get the 10 they needed was very expensive and the issue of cost was not something they had had to deal with directly before.

There was also the issue of the fragmented nature of scholarly literature. In the past they had used link resolvers and other tools; now they found themselves relying on Google, Google Scholar and PubMed.

Our customers say that their scientists stay current in their field by spending time scouring the web, emailing former colleagues for papers, or driving to university libraries. They could easily spend five to 10 hours each, every month, looking for articles. These are hidden costs. Scientists also know that this approach is not in the spirit of copyright.

How do you address this?

As a company we believe in the value that publishers bring, and in copyright law. We want to partner and support them but also to serve the unaffiliated user better. Out of this came the idea of rental. Users get to read the papers in their browser, but they can’t download them. By making that limitation, we can significantly reduce the cost.

The concept began about three years ago. As well as prototyping and user testing we’ve been working on partnership development – getting publishers on board. The notion of rental and of serving this audience was new to the industry but publishers, to their credit, are very willing to experiment. We now have well over 100 publisher partners, including Springer, Wiley, Nature Publishing Group; university presses such as Oxford University Press; and many society publishers such as IEEE.

We could show that it didn’t overlap with publishers’ existing business. We don’t restrict who can access the service – they just have to give their email and credit card details. However, less than five per cent of the email addresses supplied are .edu, .gov or .org. The majority are .co or .com. Academic researchers are likely to have access to more subscription resources anyway and often need to download PDFs so they can read them while they travel and mark them up.

How has this developed?

We want to continue to try to be innovative and meet researchers’ needs. This was the idea behind our free, five-minute preview service. For many years the publishing industry has had a free taster – called the abstract. However, users have said that the abstract wasn’t enough. We wanted to experiment with letting users go beyond the abstract and letting them preview the whole article.

Based on user interviews, we decided that five minutes was about right and, so far, we’ve had virtually no negative response to the five-minute limit. Although users would like it to be longer, five minutes is much better than nothing. We’ve seen a slight increase in traffic as a result but a huge increase in subscriptions. Our hypothesis is that, as users get to see the benefits, it will lead to more users. This a monetisation opportunity but, for us and also for publishers, it fulfils the aim of improving dissemination.

What are your thoughts on open access?

This industry is going through a lot of innovation and transformation. Open access has been developed over many years. To me it’s just another way to access something. I don’t see this replacing paying.

The competitive angle will be how much the author pays – but the risk is that, as the price goes down, there is the potential for quality to be much more mixed. The way users behave is that they are drawn to brands that they recognise – so the irony is that open access could push users towards traditional journals, either because of too much choice or fears over quality.

We have just as much open-access content in our service as paid. To deliver on the promise of convenience we want to aggregate everything. We don’t make money from it, but think that our content is richer for having it.

However, we’re not just trying to sign up every publisher. As a start-up we have limited resources, so need to be more targeted.

What changes do you predict?

In my opinion, researchers are social creatures – not like consumers behave with sharing photos, but social in terms of sharing information. That is why conferences and learned societies exist. However, this hasn’t so far really translated to the web. A lot of information that could be shared is behind paywalls and a lot is not worth sharing or can’t be. We believe our five-minute preview starts to address this. It gives real information and good control over public or private sharing.

I think the scholarly publishing industry is quite similar to how enterprise systems were 10 years ago. People said their information was too sensitive to be in the cloud, but now many enterprise solutions are cloud-based. I think this is where scholarly information will go. This will really facilitate information sharing and give much more consistent analysis of how information is accessed and used.

Interview by Siân Harris