For better or for worse, scholarly authors care about Impact Factor. While there is certainly a lot of buzz these days about “altmetrics,” the fact remains, the Impact Factor is a top consideration when authors are selecting a journal to publish in.
So the question is: what can scholarly publishers do to boost the Impact Factor of their journals? While publishing great content is certainly necessary, it’s ultimately not a sufficient strategy to optimize citation rates.
Born under mountains of information overload, we launched TrendMD in April 2014 with the humble goal of wanting to help publishers connect the right content to the right audience. To do so, we set out on a path to create a recommendation engine that looks beyond relatedness, and instead, overtime learns and adapts to your audiences behavior. By creating a tool that surfaces increasingly personalized article links, our vision was to help content producers drive more discovery of their articles, with the ultimate goal of driving citations.
But talk is cheap. And as any good group of scientists, we’ve dug into our data to see how we’re doing.
To test our hypothesis, we tracked the average monthly click-through rates on the widget for two top journals (i.e. JMIR and BMJ Open) out of our network of now over 225 journals.
According to the data, our hybrid approach of using both relatedness and collaborative filtering (i.e. utilizing both crowd and cookie level click data to generate recommendation links), yielded an average increase in engagement (as measured by click-through rates) of greater than 75% when compared to using just relatedness alone. Some items to keep in mind while reading the graph: we rolled out our collaborative filtering technology in late July, and the reason for the differences in engagement between JMIR and BMJ Open is likely due to variations in the widget placement on the two journals.
But now you’re probably asking yourself: “So what? Who cares about engagement and an increase in article views?”
When we dug deeper into our data, we found that clicked articles included in our recommendation pool received an average 29% increase in saves of those articles on reference managers (i.e. Mendeley, EndNote, etc.).
The reason why these results matter is because there is a strong correlation between the number of saves on reference managers and citations at 1 to 2 years later. Based on this, we hypothesize, that in time, the internal links on the widget may lead to an increase in the average number of citations for articles (i.e. an increase in Impact Factor).
So, to sum up - our data suggests that adding the free TrendMD recommendation widget to your journals will not only make your readers happy discovering new and interesting content, but it seems that it may also help you boost your journals’ Impact Factor.
If you want to pilot the widget on your journals, feel free contact Paul ([email protected]).
The widget is free, simple to install (3 lines of code), and the results are all readily measurable.