REPRISE: How I Hosted the Tangled Bank ..

... and what I did to make it a success..

NOTE: this is reposted here from my original blogger site at the urging of my friend and colleague, Bora. Thanks Bora for noticing and remembering this piece!

When I hosted the 23rd issue of the Tangled Bank (TB23) on 9 March 2005, I didn't know what to expect because I had never hosted a blog carnival before and in fact, I had only recently figured out what a blog carnival actually is. Nevertheless, my primary goals were to have fun and to avoid the pitfalls experienced by previous hosts. One of the worst problems that recent issues of TB experienced was technological. The previous TB host's anti-spam software trashed most submissions, so few stories were received and linked for issue 22. I was determined to circumvent the ravages of anti-spam software, so I decided to email acknowledgments to each author so they knew their submissions had been received. Additionally, because I had read, thought about and observed TB during the previous couple months, I took the liberty of developing a few minor goals of my own for "my issue", goals that (I thought) nicely complimented the original, more encompassing purpose of TB.

As I stated, my main objective was to have fun yet still make it a success. First, to prevent TB23 from becoming a chore, I broke it up into smaller pieces that could easily be accomplished each day and I also made sure that my working and master files were properly backed up. To do this, I made two TB23 files; one was on my blogger account and the other -- the master copy -- resided on my computer desktop. Early every morning, I worked on TB, reading and summarizing each article that had been submitted during the previous 24-hour period of time. This work consumed between 15 minutes to 2 hours each morning and was saved on my computer desktop file. When Blogger was functional, I then copied the entire desktop file to my blogger TB23 file and saved it.

This duplicate system of using a desktop "master" file only consumed a few extra minutes each day but prevented damage due to Blogger crashes (always a possibility), and copying the entire file to my Blogger account allowed me to work on TB23 anywhere and any time that I had internet access while simultaneously safeguarding this file from corruption due to computer crashes ("my" borrowed computer is very unstable). For those of you who are curious, within the two-week period of time that I worked on "my issue" of TB, I experienced one Blogger crash (the day before TB23 was to be published) and two major computer crashes, any one of which would have spelled disaster for this project.

My time commitment abruptly increased on Monday, several days before TB23 was to be published. During the three days prior to publishing TB23, I was consistently investing two and sometimes as much as three hours reading and summarizing articles. I initially tried to work additional time late in the evening on TB, but found that I was not as coherent as I wished the following morning, so I instead shifted my entire time commitment to the morning. This worked out better.

The morning that I posted TB23, I finished writing the lead opening paragraphs, made final edits to the master draft, rearranged the topic categories (again!) so the topics "flowed" better, and I tried to write logical transitions from one article to the next to keep my readers' interest. I double-checked all internal links on my authors' articles and notified them of any that were broken (I found only one broken link and the author was able to fix it within a few hours).

To increase the number and variety of articles submitted, I published two appeals for submissions to my blog; one was posted two weeks before the publication date and the second was posted one week prior to publication. Then, regardless of where in the city I was located, I could check my email account for new submissions and add them to my Blogger file, beginning the very day that TB22 was published. (You might think this was premature, but my first three submissions arrived that very morning, in fact!).

Because I strongly disliked wondering if my own submissions to previous issues of TB (and other blog carnivals) had been received, and because I suspected that other authors felt the same way, I acknowledged each submission as I received it. As each story was sent to me, I replied to the author in 24 hours or less (and cc'd this acknowledgment to PZ when he had forwarded the original message to me) and then copied the text of the original email into my desktop master file (which was usually copied to blogger later) so I could work on it later. I also saved all TB emails to a special file on my email account, so all would not be lost if a "supercrash" somehow managed to destroy both TB files.

As I mentioned earlier, I developed several minor goals of my own for this issue of TB. These goals were to encourage contributions from female "bloggers", to seek contributions from "new voices" and, because blogging is a global enterprise, I wanted to foster a more international flavor by encouraging contributions from "non-Americans". To this end, I approached some of my own contacts in the blogosphere for particular articles that they gladly contributed to "my" issue, and one TB contributor unexpectedly helped by telling me about some "interesting blogs" that I should check out. These "interesting blogs" all contained material that were ultimately included in TB23, and the authors were flattered to be approached by me (some of whom have since sent ecards, personal emails and other tokens of their appreciation). I am pleased to say that I think I achieved all of my goals.

On the publication date after all my work was finished, I copied the entire completed master file to my Blogger account and .. held my breath while the wheels of Blogger publishing churned .. somewhat clunkily .. and then .. there it was! After quickly checking the document for formatting issues that needed fixing, I sent email to all contributors and included the URL so they all could link to TB23 from their blogs (14 of them did include links).

On Wednesday, the first day that TB23 was publicly accessible, my hit meter registered 574 hits (my normal number of hits per day was between 90-150), so this was roughly a three- to four-fold increase in visitors to my site. On the following day, Thursday, MSNBC somehow discovered us and linked to TB23 in a story describing the power of blogs and bloggers. The result was a respectable 301 visitors for that day, approximately one-third to one-half of which were from that one story. In fact, this one MSNBC story resulted in a steady stream of 20-50 readers each day for more than a week and, as of today, I still receive 2-10 visitors each day from this one reference.

Then on Friday afternoon, InstaPundit finally honored my repeated requests and linked to TB23 in the early afternoon (EST). Within minutes, my site meter registered a dramatic jump in visitors: I saw a greater than ten-fold increase in hits, from 30 per hour to more than 300 per hour. The peak number of visits from the InstaPundit link were 421 per hour and total visits for that Friday alone were 2,107 -- a number that diminished to "only" 636 on Saturday. Don't forget that the number of Saturday visits was still higher than the total number of hits that TB23 received on Wednesday, which is traditionally the single day with the highest number of visitors. Surprisingly, visits from InstaPundit dropped off more rapidly than those from the MSNBC story and I rarely received more than one or two hits from this reference five days after it was published.

I hope this little document can help you make the decision about whether to host TB, and I hope it gives you some ideas for how to organize your time and efforts if you do host it. In short, the entire TB experience was wonderful and I highly recommend it. I read a lot of good science writing and I happily seized this rare opportunity to "meet" all of the contributors through email, to approach and add new contributors to our growing group, to "meet" new readers who subsequently flooded my email box, and I managed to make the contributors happy with additional traffic to their sites as well as with the end product, the TB23 index (or so I think). Hosting TB23 also allowed me to indulge several of my personal goals; to share good science, medical and nature writing with the public and to encourage scientists, medical doctors and others to continue writing by providing a venue for them to share their essays with the public. I also had the added -- and unexpected -- benefit of building stronger personal connections with all of the contributors. In short, TB23 was one of the most positive and pleasant experiences I've had in the past 8 months and I would happily (eagerly) host it again.

NOTE: Originally published here after I'd successfully hosted my first blog carnival ever, and my first issue of Tangled Bank. I am thinking I need to update the traffic numbers as they apply to what I get from this site -- and if you are wondering if I am interested to host more blog carnivals, the answer is a resounding yes! It's fun!