Editor’s Note: this post is by Lettie Conrad. Lettie is the Executive Product Analysis Manager at SAGE. Lettie leads a group of specialists who expand SAGE’s capacity for digital product development innovations. She is instrumental in launching user-centered web and mobile products, driving research and analysis that enable evidence-based product management.
Today’s scholars navigate a tangled web of databases, search engines and applications throughout their academic endeavors. Researchers often report that their information seeking and retrieval processes are fraught with detours, confusing sign posts, and frustrating authentication checkpoints – as one PhD scholar demonstrated last year at the London Book Fair (starting on slide #4). This brings to mind an image generated by a study presented to ACRL’s annual meeting, which illuminates the knotted network of starting points in academic research. To truly engage with today’s academic community, the scholarly communications industry must be willing to open up multiple avenues into our resources.
For some scholars, it’s a matter of preference – “Google is familiar territory.” And others come to prefer a leading subject database in their field of study – “with PubMed, everything I need is archived…it’s the thing to use.” Many researchers say they mainly use their library discovery tool for “known item” searches, when they’re looking for the full-text of an article or monograph. It’s become clear that the pathways into academic material are diverse and the researcher experiences (RX) are quite variable. The fact is: Scholars today are becoming experts at traversing an ever-changing system of onramps onto the scholarly information superhighway.
This RX reality means that we need to support the way research is really done (not the way we wish it would be done). With the rise of Google and the advent of pre-indexed library discovery services, SAGE has adopted a strategy that aims for wide participation in a variety of discovery channels. For a 50-year old publisher, this is a bag of new tricks. An increasingly liberal methodology in metadata distribution has evolved slowly over the last 10 years, as we learn what works and what doesn’t work for our readers and library customers.
Over time, we’ve come to terms with the fact that our web properties are not destination sites, but that we must take a multi-faceted approach to driving our bibliographic data into as many relevant channels as we can. Today, SAGE mobilizes our published metadata for delivery to indexers and aggregators far and wide – to the point that, as of today, we see our metadata as quite promiscuous in nature, well-known in many ports of call across the global scholarly ecosystem. This is made possible by a committed team of experts, working with an XML-based content repository; a system that replaced many hours of manual file transfers and has been customized to automate distribution of metadata to dozens of indexers within seconds.
We now have many routines in place to support broad discovery of SAGE content, from MARC records to classic SEO. Not to say we’ve got it all figured out – on the contrary, the quality and accuracy of our data and our distribution capabilities are works-in-progress. We believe strongly in our allegiance to prevailing information and metadata standards, such as our recent announcement about our conformance with the Open Discovery Initiative. And we remain open to new opportunities – like Kudos, ScienceScape, Browzine, and the Wikipedia Library program.
From the mainstream to the specialized, we aim to make our content visible in a range of search tools, databases, applications and social networks, where they’re relevant for or popular with SAGE readers. Of course, there are the usual caution signs to watch out for and we debate the pros / cons of experiments as they arise. We keep in mind that despite our best efforts, there will always be web-design limitations to making SAGE content stand out in the crowd. And, through R&D projects, we’ve found that some content channels are not relevant for our users and some indexing experiments turn out to be dead-ends.
We like to start small and build on what we learn through doing. For instance, reference discovery is a particular challenge – so when we find success with one partner, we do what we can to ramp up indexing and partnership opportunities where we see evidence that they effectively support today’s researcher experience. On the flip side, we’ll cancel a new pilot that initiative isn’t driving sufficient traffic to warrant the expense, or we learn that other channels are more popular with our readers. The goal is to integrate those lessons into our overall strategy and keep moving forward, to find new avenues for our metadata to support the scholarly enterprise.
Some new discovery vehicles take quite a long time to demonstrate their value, especially with seasonal academic cycles creating “twin peaks” of activity in spring and fall. Some opportunities (like subject-specialized pathways) have intangible, intrinsic benefits that aren’t reflected in sales data or usage statistics. We’re finding that not every new pilot initiative is a hit with our readers and customers – and some prove too great a challenge to our existing ways of doing business.
Publishers should draft their discoverability strategies in pencil and be prepared to redraw their roadmaps and adapt to the search and browse preferences and research habits of their readers. For the sake of good usage ratios, to optimize the RX of our content, and to succeed in this brave new world, SAGE has found that an expansive approach to discoverability best serves our authors – and the scholars that cite them, the faculty who assign them, and the students that read them.
4 Thoughts on "Guest Post: Lettie Conrad on Metadata Promiscuity and the Researcher Experience"
Sounds like fun, Lettie. I argue that the science of search is one of the grand challenges of our day. In that regard, you say “…we’ve found that some content channels are not relevant for our users and some indexing experiments turn out to be dead-ends.” Can you tell us more about some of that? Negative results are very important in this context.
Totally agree, David, understanding what does not work for discovery and search are just as, if not more, important than what does work. We’ve tried indexing with some mainstream avenues that target non-academic “knowledge workers” with an “iTunes” sales models that didn’t have as much traction as we’d hoped. I’m wary to “name names” in this context, as something that didn’t strike gold with our researchers last year might become a new horizon for us next year!