Going beyond the search bar: database publishing and how database usage can be measured

At Springer Nature we believe our mission is to support researchers in advancing discovery, and we strive to do this by publishing around 300,000 journal articles and over 13,000 academic books every year and embarking on initiatives, projects and innovations to ensure the discoverability and ease of access of this research. We are always looking for new ways to serve our community and ensure that we are both supporting, enhancing and delivering what our communities need.

In an advanced digital age, researchers must constantly adapt to new and innovative ways to conduct and access research quickly and effectively. One way publishers can support researchers is through the creation of innovative, dynamic databases. 

The growth of data products now far exceeds that of journals and books, and they are a complimentary resource, supporting journals and books by offering discovery, analytics and data integration tools; transforming the ways in which researchers search for information and enhancing the research journey.

This is because databases are able to address specific challenges the research community face: 

• Users need easy and intuitive ways to search through our extensive content. 
• Researchers must be able to interrogate data to find highly specific pieces of information.
• Information is often available in isolation, and researchers need to find a way of connecting these ‘islands’ through meaningful relationships.

It is perhaps easiest to think about the value of databases in the context of a researcher’s journey and the challenges they may face: a young post-doc is struggling to find out why her experiment is not working; a PhD student working on drug discovery and wants to know what related drugs are currently in clinical trials; a scientist working in industry needs the melting point of a three-metal alloy under high pressure; or a materials scientist is looking for the dimension of a carbon nanotube.

Not so long ago, these researchers would have had to search for the information they needed by looking through print journals and books. Being able to access these resources digitally opened doors to a completely new way of providing content to readers. 

In the age of databases, the young post-doc could search Springer Nature Experiments, the PhD student can see the whole landscape of relevant drugs on AdisInsight,  a quick search in Springer Materials would give the industrial scientist the melting point he needs, and the materials scientist could get the dimension of the carbon nanotube from Nano.

At their core, therefore, a database’s value is in its ability to organise scientific information so that it is structured and searchable in a way that is specific to the content. There is no “one size fits all” approach. Developing a database requires knowledge, filters, concept recognition, information extraction, relations, categorizations, classifications and more.

One of the challenges with data products, however, centres around getting an accurate understanding of its usage and therefore clear insight into its impact and value for the user. For electronic books and journals, one way we can do this is to simply look at the number of times an article or chapter has been downloaded. A higher number of downloads generally means that the publication has been used – and hopefully read – more often by more people. This type of measurement doesn’t exist for databases and so when improving, developing, assessing their value –  the major challenge is being able to accurately measure and understand how frequently and in what volume database content is being used. 

With databases, users do not necessarily download content; sometimes it’s enough to read the content on the screen or to click a button. 

We can also not measure the amount of time that a researcher spends using a database, because ideally, they should be able to find their answer quickly and efficiently. If they are spending a lot of time searching, this could indicate that we haven’t done our job properly and functionality isn’t a clean as it should be. On the other hand, this is not as simple as that, because it really depends on which data product we are talking about and who our user is. 

Digital solutions are evolving and the metrics used to measure our databases will continue to be redefined, but we need to stay on top of and aware of the changes. For example, how often users return to our sites and what functionality they interact with is information that is vital for us to understand how satisfied and engaged the users are with our products. But if we do our job too well, and make our search function too smart, we might start seeing a decrease in both, and this could be wrongly interpreted. We need to continuously monitor our metrics, learn what they mean and what impact functionality changes has – this is our challenge for the coming years.  

Where to next?

We are passionate about creating great databases that researchers can use to find the information they need. We currently have over 200,000 database users, so, we must be doing something right. But the market continues to change and develop – we always need to be looking ahead at the next development that will further improve the databases that we offer.

So what’s next for us? We are already using Artificial intelligence (AI) and machine learning (ML) to make the discovery more intuitive and dissolve the information “islands”. But we are just at the beginning and our vision is to expand that even further.

We need to provide solutions to specific problems that face the communities that we serve. As trusted accelerators of knowledge, we have a commitment to stand alongside our community, supporting and making change happen in the format and with the resources that they need. As database enthusiasts we need to make sure our users come to our sites, obtain what they need and leave with a smile on their faces.