Tilting the balance back towards libraries

Share this on social media:

Jason Priem tells of his hopes for a ‘long-overdue’ change in academic publishing

Tell us a little about your background and qualifications?

I was a middle-school teacher for five years, teaching language arts, social studies, and media. As a teacher, I started to realise how big an influence the Web was going to have on learning and knowledge, so I taught myself to code and dove in! 

I worked on an information science PhD for four years at the University of North Carolina – Chapel Hill. My work there got me interested in how social media activity might open a new window on scholarly impact, leading me to coin the word ‘altmetrics’ and co-author an ‘Altmetrics Manifesto’.

As an academic, I learned about the open access and open science movements, and got really excited about that... I felt like I had to build something to help!

What is the history of Our Research?

Heather and met one another when I peer-reviewed one of her papers; later we met in person as part of an open science hackathon. The hackathon ended in the evening, but we went out in the hall of the hotel and kept working all through the night. I remember being surprised when the hall started getting crowded by waiters serving breakfast the next morning! 

That hackathon project went on to become Impactstory Profiles, our first big website. From there, we built Unpaywall, a free and open index of all the world’s OA articles. Many libraries asked us to build a subscription analysis tool using Unpaywall data, so we did. That became Unsub, which launched (as ‘Unpaywall Journals’) in November of 2019. And now hopefully we still have a lot more history left to write!

Our nonprofit mission has remained the same throughout: we build tools to help further the progress of open science, because we believe research progress is more efficient and effective when it's open.

In the UK, Jisc has just signed up to your Unsub service. Can you explain how it works?

Certainly. Unsub is an analytic dashboard that helps academic librarians cancel their subscriptions to ‘big deals’. These ‘big deals’ bundle up thousands of toll-access journals from a single publisher into a single, massive subscription. They have become the central pillar of the entire toll-access publishing apparatus, because they are highly profitable – a single deal for a US R1 university is generally several million dollars – and libraries have increasingly felt ‘locked in’ to these deals due to a perceived lack of alternatives.

The growing prevalence of open access (OA), however, offers a solution to the captivity of libraries at the hands of these big deals. Because much of the content of the big deal is now available as OA, there is now  a smoother ‘off-ramp’ to cancellation than ever before. That is, after cancellation, faculty can still access a large percentage of the relevant literature. Librarians are increasingly aware of this. 

However, there has been no way to quantify the exact percentage of post-cancellation access that OA is able to provide. Although a growing number libraries have cancelled, the lack of hard numbers has made most librarians unwilling to upset the status quo.

This presents a compelling opportunity for us as OA advocates: by helping libraries quantify the alternatives to toll-access publishing, we can empower librarians to cancel multi-million dollar big deals. This in turn will begin to turn off the faucet of money flowing from universities to toll-access publishing houses. In short: by helping libraries cancel big deals, we can make toll-access publishing less profitable, and accelerate the transition toward universal OA.

Unsub is designed to do this. It creates a set of forecasts, customised to a given library, that the library can use to understand the impacts of cancellation. This requires the creation of a usage model for a given scenario. The usage model incorporates library-specific data for citation, faculty authorship, campus downloads, and pricing information.  It also includes global data on interlibrary loan rates, open access, disciplinary readership patterns, and many other factors. Libraries can copy, tune, and customise this model in many ways, creating a reliable and objective plan for their future without toll-access big deals. This post-big-deal future is generally quite a bit rosier than they expected...this in turn gives them the confidence to cancel.

What led you to develop the product?

 Well, it would be nice to say we came up with the idea because we were very clever...but I’m afraid that’s not it. In reality, libraries just kept asking us to build it, and finally we did!

As I mentioned, many libraries are under intense and growing budget pressure (which the pandemic has not improved). And for the last 20 years, the balance of power in the relationship between libraries and publisher has tilted ever more toward the publishers. So there is a lot of demand out there for a tool like Unsub that can help tilt the balance back toward libraries. I think that’s a long-overdue change.

The tool offers three main advantages for libraries, over their current workflows:

• Unsub is more comprehensive. Our model accounts for the effect of OA (green, hybrid, bronze, and delayed), COUNTER downloads, previously-purchased backfile, interlibrary loan, document delivery, faculty citation and authorship patterns, journal readership decay curves, and then shows how that all affects fulfillment rates and costs. It’s just a more complete picture.

• Unsub forecasts the future. Instead of just looking at the current state of a library’s collection, Unsub uses a forecasting model (trained on millions of data points) to simulate the future. Users can experiment to see how their plans will affect costs and fulfillment rates for the next five years--it’s kind of like playing SimCity with a serials collection.

• Unsub saves time. For most users, it takes just hours to set up an Unsub profile; they just upload your COUNTER report, perpetual access status, and price lists. After that, it’s all gravy: Unsub replaces annoying and time-consuming spreadsheet-juggling with a simple UI. By lightening the analytics workload, Unsub gives institutions the ability to take a much deeper look into their options. We’ve had many users tell us that it feels great to walk into negotiations more prepared and data-informed than they’ve ever been.

It’s now been about 18 months since we launched, and we’re used by roughly 400 libraries worldwide, including a lot of the world’s most prominent institutions. We think that’s a credit to our terrific user community, who has done a lot to spread the word for us.

What are your hopes for the future of scholarly communications?

That’s a great question, but a tough question to answer briefly! In fact, I’ve written at some length about this elsewhere, and I still hold to most of that.

Our goal at Our Research is to accelerate the transition toward universal Open Science. That means we want a world where the default is set to open for all research products, including papers, preprints, datasets, source code, protocols, and more.

This world isn’t one of openness for its own sake, but one where all these delightful open products of research can be processed, remixed, distributed, summarised, annotated, text-mined, and used by an open ecosystem of automated tools. 

Scholars invented the Web, for scholarship. But 25 years later, we still don’t really use it! Or when we do, we use it in simple, unimaginative ways. That’s largely because we just don’t have access to scholarship on the open web – it’s siloed and balkanised and paywalled by a clumsy swarm of for-profit publishers (as well as, increasingly, for-profit social networks like ResearchGate).

The world of ideas is a singular one – every idea can be viewed in the context of any other idea. The scholarly communication system is the humble substrate of this process, the infrastructure, the subway where all these ideas and data can ride around and get to know one another. Let’s make that subway as well-connected and cheap and easy-to-use as we possibly can. Today’s world is asking a lot from the research community. We owe it to them to build on the best system we can.

Do you have any hobbies or interests you want to tell us about?

I really enjoy what we are doing at Our Research... I’m very fortunate that it still feels like a fun hobby to me, as well as a job. So I spend a lot of time on that. I also like to hit the gym, play volleyball and jiu jitsu (when there’s not a global pandemic on, anyway), relax at the beach, play Dungeons & Dragons, hang out with friends, and of course waste time in front of Netflix :)

Interview by Tim Gillett