We recently completed a set of crossover studies over a 6-week period to determine TrendMD’s effect on weekly pageviews for a group of 1000 randomly selected articles published in BMJ Open. Overall we found that TrendMD increased weekly pageviews by 28% relative to baseline weekly traffic for the group of articles.

Here’s what we did:

First, we randomly selected 1000 articles from the 3,752 published to date by BMJ Open.

First week: We determined baseline traffic (control) for the group of articles by disabling links to the articles from appearing in the TrendMD widget results on both BMJ Group sites (internal recommendations from articles published in the 10 other BMJ Group sites at the time of study) and from third-party sites in the TrendMD network (external recommendations). Baseline traffic to the group of articles during the first week was 3,862 pageviews.

Second week: Links to the 1000 articles were enabled from both the publisher’s own content and from the TrendMD network, but no marketing budget was assigned to promote articles across the network. Articles were promoted solely using credits earned by the display of 3rd party articles on the BMJ Open site. An additional 348 pageviews were generated from links appearing on other BMJ group journals (internal recommendations) within the TrendMD widget on 10 other BMJ Group journals. The BMJ Open articles also earned 39 traffic credits from 78 clicks on affiliate links displayed in widgets installed across the 1000 articles pages. An additional 39 pageviews were thereby acquired from clicks on affiliate links to the BMJ Open articles displayed in widget results on third-party sites (external-earned recommendations). Taken together, TrendMD used without any additional marketing spend, increased weekly pageviews for the group of articles by 10.4%, relative to the control week.

Third week: We repeated the control conditions; 3,859 pageviews were recorded, very similar to week one, indicating that background usage levels remained steady.

Fourth week: We added $250 in traffic credits ($1 USD per credit) for promotion of links to the 1000 articles across the TrendMD network. During the fourth week, 347 pageviews were generated via internal recommendations, and an additional 292 pageviews were acquired from external recommendations (250 paid, 42 earned). The addition of a $250 budget enabled a 16.3% increase in week-to-week pageviews.

Fifth week: Another control week, in which 3,868 pageviews were recorded.

Sixth week: BMJ Group rolled out the widget to 4 more sites and we increased their equivalent budget to $500 in traffic credits. Both changes lead to a 28.5% increase in weekly pageviews for the group of articles, relative to control. An additional 558 pageviews were generated from internal recommendations, and an additional 546 pageviews were acquired from external recommendations (500 paid, 92 earned). The TrendMD network was able to fully utilize the marketing budget to deliver targeted traffic - i.e. the ability to deliver traffic was limited only by budget, not by TrendMD’s network.

Taken together, this crossover study indicates that both paid and unpaid distribution of article links over the TrendMD network can be an effective way to increase weekly pageviews of a group of scholarly articles.

Next up, we’re measuring effects of TrendMD on Altmetric scores. Stay tuned!

Feel free to contact us for more information.