In 2010, a few of my staff members began pressuring me to start a blog. Although my friends and colleagues will attest that I'm not a shy or soft-spoken individual, I was, to put it mildly, reticent. Blogging was relatively new territory for government officials at the time. There were some outstanding examples of blogs by science-oriented government offices, such as the White House Office of Science and Technology Policy, which by then had been running for a couple of years (www.whitehouse.gov/administration/eop/ostp/blog). Jeremy Berg, former director of the National Institute of General Medical Sciences, had also successfully established a following on the blog Feedback Loop (https://loop.nigms.nih.gov), so I knew that it could be done well. But how would people respond to a National Institutes of Health (NIH) blog on extramural research policy? And how would it fit in to our work of supporting biomedical research?

My staff knew that I wanted transparency to be a centrepiece of my tenure at the NIH, and felt that the biomedical research community would find it refreshing to hear more from NIH management. Revealing how we do our business as we are developing policies and processes, and presenting data and information that shed light on the impact of our decisions, seemed like the right thing to do.

Credit: ILLUSTRATION BY DAVID PARKINS

On 19 January 2011, I launched the Rock Talk blog (http://nexus.od.nih.gov/all/rock-talk). Two years later, I'm delighted that I did (even though I have taken heat for using too many exclamation marks!). As of 2 January 2013, I've written 107 blog posts, averaging about 40,000 page views a month. A steady stream of comments keeps me on my toes (see 'Top ten hits'). For the first time, the NIH was presenting data from all of its institutes and centres, and was sharing its analysis of funding trends on a forum where scientists could give their own opinions and reactions directly.

Table 1 Top ten hits: Ranked by number of individual page views, the ten most popular entries on the NIH Rock Talk blog generated varying numbers of comments.
Full size table

My first data-sharing post looked at trends in funding at various types of research institution — showing, for instance, that the majority of NIH funds go to medical schools, and that distributions to for-profit institutions have increased since the 1980s. Within two weeks, the post generated nearly 90 comments, about everything from the ethics of training non-US scientists to the importance of including more MD-trained investigators.

A sense of community

Conversation is an integral part of the policy-making process.

We learned a lot about our constituents' interests and needs through the blog, and we have been able to highlight behind-the-scenes data, actively engage the community in policy-making and provide insight into our decisions. Without public input, effective and impactful policy cannot be created. My blogging experience has convinced me that using social-media platforms is one effective way for science-funding agencies to successfully support research.

The blog is not the official vehicle for communicating policy changes. We have the NIH Guide for Grants and Contracts for that. And we have official channels for soliciting public feedback — through requests for information in the NIH Guide, for example. But the blog allows me to extend that conversation to people worldwide, many of whom I would not be able to reach in other ways. People are sometimes concerned that offering a dissenting opinion to NIH officials might affect their chances of getting funding. Although that is absolutely not the case, one advantage the blog does have over some other channels is that it allows people to remain anonymous if they wish.

Another attribute of the blog — which I think should be mirrored by most government blogs — is how we communicate current events in real time. For example, when Hurricane Sandy hit the US east coast in October 2012, we posted advice on what awardees should do if their labs were affected by the storm, and highlighted relevant resources and guidance for dealing with natural disasters.

I am often asked whether I really read and respond to comments in person. I do check them every day, on average; some I answer myself and some with the assistance of my blog team (made up of four staff who help regularly, plus five or so who contribute when needed). At times, the criticism of the NIH or even of me is harsh, and can feel personal. But I understand. In this time of tightening budgets, the topic of funding is very personal, and I understand the commenters' frustrations. The blog is moderated and we have had to screen out a very few comments that are off-topic, constitute a personal attack on another commenter or discuss the details of individual grants or applications (see the commenting policy at go.nature.com/ipwxqs).

The blog post that has received the most comments to date, 'How Do You Think We Should Manage Science in Fiscally Challenging Times?', was also one of the shortest. In fewer than 300 words, it asked scientists for their thoughts on how best to manage the NIH during times of austerity. This level of transparency into our thinking about future funding strategies was a novelty. The post also linked to interactive data charts that allowed readers to manipulate variables to see how their proposed changes to the size and numbers of grants we fund would affect success rates. We received more than 250 comments, including suggestions to take a closer look at how much funding was being given to individual scientists. We reviewed the suggestions and instituted our new Special Council Review policy to provide additional consideration of applications from principal investigators with existing NIH research project grants of US$1 million or more in direct costs (see J. M. Berg Nature 489, 203; 2012).

Blog posts containing data have clearly been the most popular (see table). Another common theme of my posts is debunking urban myths about NIH-supported research. Ideas for these posts come from blog readers' comments, grantee e-mails, discussions with attendees at research conferences and more. For example, in 'Myth Busting: Number of Grants per Investigator' we showed that the myth that many principal investigators continuously add to their bounty by piling on more and more NIH awards is untrue. The data we provided showed that the investigators who received the top 20% of funding in 2009 averaged only 2.2 grants each. (This example also illustrates how readers interact with the blog — they noticed that we had posted the wrong information in one of our graphs, which led to a follow-up post correcting the information and elaborating on the data we presented.)

Blogging is a lot of work, but it is also fun and enlightening. It would not be possible without the help of my Rock Talk blog and data teams, who provide statistical analyses, help me to decide on which topics to cover and moderate and respond to comments.

I encourage others in science policy to go into blogging with their eyes open, recognizing the level of effort required to care for and maintain a blog. Potential bloggers should be aware of the long-term commitment needed to determine which issues are blog-worthy and of interest to the community. Rock Talk has sometimes covered topics that reveal NIH idiosyncrasies. But I think that is what makes the blog real and helps us to connect with the community.

If you are not a reader of Rock Talk, I hope you will check it out. If you are, hopefully you know by now that I welcome your ideas, through either commenting on the blog or even starting a blog of your own. Conversation is an integral part of the policy-making process, and helps the biomedical research enterprise to keep on rockin'.