FluPortal.org

A Review of FluPortal and Recommendations for the Future

View the the full report or download a .pdf version.
1. Summary
2. Technology and Design
3. Blogging and Curation
4. Outreach and Communication
5. Appendix

FluPortal Blogging and Curation

by Katherine Bidwell

Overview

When I joined the FluPortal team in mid-September 2009, Rekha and Josh had done the hard work of figuring out the structure of the site, had built up many of the resources on it, and had written some initial blog posts. They asked me to take charge of the blog, add actively to the Delicious bookmarks, and keep an editorial eye on the site as a whole.

The Blog

The Blog’s Initial Mandate
Rekha and Josh had already worked on initial ideas about the blog’s mission. Here’s the mandate they proposed in September:

Audience: The purpose of the blog (like that of the rest of the site) was to help pubmedia report on H1N1. So the audience was very specifically pubmedia reporters, news directors, and web managers — not the general public.

Topics for blog posts: The goal was to present five main things:

• Important government updates
• Interesting public-media coverage
• Web tools that stations could use on H1N1 pages
• Major H1N1 news stories
• Any FluPortal news (webinar dates, site improvements, etc.)

Minimal original reporting: The blog wasn’t a place for original reporting. If we interviewed people for posts, the point was to offer ideas — teases, really — that journalists could then report on in-depth themselves.

Posting frequency: The goal was to post one or (max) two posts a day so as not to inundate readers.

Guest posts: The premise was that we’d do most of the blogging ourselves but that we’d also recruit occasional “guest posts” from pubmedia staff and public-health or medical experts. (Details below.)

How the Blog Evolved
Josh and I were the primary bloggers, and Rekha and Ken contributed several posts as their schedules permitted.

Over time, we incorporated the initial goals and added a few other elements. This is how I’d now describe the blog: We aimed to be a source of inspiration for reporters looking for story leads and story angles. We also offered tech tips for reporters and web managers — tools they could use in reporting, in building an H1N1 web page, and in getting their stories and other information out to the public.

Here’s what we added to the initial five-point list of blogworthy topics:

• Photos and video from social media that reporters could use to illustrate stories or use as story leads
• (As H1N1 tapered off over the holidays:) generalized resources on crisis reporting that could be applied to H1N1 or other future crises

We tried to keep the tone of the blog informative, neutral, and not too formal.

How We Put Blog Posts Together: The Nuts and Bolts
How I Found Ideas for Posts: Every morning, I’d scour the news. These were my main news sources (it would be easy to draw up a comparable list for any future “crisis portal”):

• A Google News search for “swine flu h1n1″
• A Google search for pubmedia stories (until the DayLife widget launched at the end of January)
• NPR’s “Swine Flu: The Second Wave” page
• NPR’s health blog
• Flu.gov’s Facebook page
• Flu.gov’s Twitter stream
• Flu.gov’s newsroom
• H1N1 pages of various newspapers
• Various flu bloggers

Rekha, Josh, and Ken were also very helpful in forwarding many ideas and contacts they ran across during their own FluPortal work.

I also regularly scanned social-media sites like Flickr, Vimeo, and YouTube for clever user-generated content. (Details in FluPortal’s social-media guide.)

Illustrations: We aimed — almost without exception — to illustrate each post with some kind of photograph, logo, or video to give it color.

If a post was about an initiative from a pubmedia station, we included the station logo and, sometimes, a headshot of a particular staff member.

Example:
“Building an H1N1 Web Page with Limited Resources”

If a post was about a photograph or video, we maximized the photo/video to fit the width of the blog. This meant it was generally around 500 pixels wide, which gave it as much visibility as possible.

Examples:
“Photos You Can Use: More H1N1 Graffiti”
“Pictures of the 1976 Swine-Flu Outbreak”

For most other posts, we used creative-commons-licensed photos from Flickr that added visual punch and, if possible, humor or a small editorial twist. We made these “illustration” photos 300 pixels wide (or 300 tall if they were vertical photos) and wrapped text around them.

Examples:
“Shifts in the H1N1 Story”
“International Info on H1N1”

Attribution and links: We made sure to attribute photos and videos with links and in a stylistically consistent way. Most of the images we posted were creative-commons licensed — so we indicated that they were “cc” and that they were freely usable by journalists/stations. (A creative-commons license allows you to use photos/videos/writing/etc. freely with attribution.)

Example of one of our typical photo attributions (showing content and style):
[Orin Zebest / cc (usable on your site) / Flickr]
See here how it appears in the post.

We used links throughout our posts, too, so that journalists could click through quickly to reports, news stories, studies, etc. if they were interested. We also used links as a form of footnotes — to back up statements we made.

Editing: We often ran post drafts by another member of the FluPortal team — especially if we felt a critical eye could help with tone or delicate editorial questions.

If a post contained quotes from an interview, we typically asked the source to check them. Our goal here — as we weren’t practicing investigative journalism — was to make sure we represented people’s views in ways they approved.

Updates: Occasionally we updated posts if we’d forgotten something, had made a mistake, or if new relevant info emerged. We always indicated that the update was an update (good blogging practice).

Example:
“Google’s Flu-Shot Locator”

Screenshot from post showing an update

Guest posts: For guest posts, we introduced each writer very briefly — in italics — and then gave the rest of the post over to their words.

Examples:
“WDET’s Ellcessor on Community Health Awareness”
“MPR’s Lorna Benson: Reporting on H1N1”

Searching the blog: One problem with a blog format is that, over time, older posts are buried and become hard to find. So we added month-based archives and eight “categories” to the sidebar — to help readers find posts. (The categories were: H1N1 data, ideas for covering H1N1, on-air, online, pubmedia coverage, social media, vaccine, and web tools.)

We also added a site-wide search box in November. This pulled results from the blog as well as the rest of the site. As of the end of February, we counted 121 unique searches (with 179 pageviews). Judging by the search terms, it looks like the FluPortal team made the majority of those 121 searches. (In other words: we frequently used the search box ourselves to locate old blog posts.)

We also made an effort, on the blog, to link back to older posts where relevant. And certain “evergreen” posts we featured more permanently on, say, the “Reporter and Program Resources” page.

Examples of “evergreen” posts linked to from the “Resources” page:
“Terminology: H1N1 v. Swine Flu”
“H1N1 v. Swine Flu: The Rematch”

How We Approached Each Type of Blog Post
1. Important government updates: We offered links to government press releases, transcripts, audio, video, etc. We also tried to pull particularly interesting quotes or ideas — to save busy reporters some time (as some of the government material was pretty dense and time-consuming to wade through).

Examples:
“Sebelius: Assessing the Federal Response to H1N1”
“WHO: We Didn’t Hype the H1N1 Pandemic”
“National Influenza Vaccine Week: 10-16 Jan”

2. Interesting public-media coverage: We avoided just pointing to coverage and instead tried to highlight informative aspects of whatever we featured — the hows and whys of it.

Examples:
“WILL’s Lessons Learned from H1N1 Coverage”
“Building an H1N1 Page with Limited Resources”
“How MPR Has Approached Reporting on H1N1”
“Joe Neel: on NPR’s H1N1 Coverage”
“A Global View of Flu from PRI’s The World”
“WFCR’s Regional Special on H1N1”
“How WBUR’s Sacha Pfeiffer Reports on H1N1”

3. Web tools that stations could use on H1N1 pages: This was largely Josh’s domain. As he added resources to the “For Station Resources” page, he typically also featured them in a blog posts.

Examples:
“New Widget Displays PubMedia H1N1 Coverage”
“Use CoveritLive for Real-Time Conversations”

4. News stories: Of the initial five goals for post topics, the “major H1N1 news stories” category was the only one that didn’t translate easily to the blog. Here’s why: Most major H1N1 news was already being reported extensively by mainstream and public media. It made no sense to waste a day’s post on information that journalists would already have in spades.

So here’s how I approached heavily-reported news: I bookmarked some of the better reporting in Delicious. (See section on Delicious below.) But I rarely blogged about the pieces.

Instead, for newsy blog posts, I tried to highlight underreported studies, smaller original angles, nuggets from government releases that might have gone unnoticed, clarifications of the confusing and ever-shifting H1N1 data, and unusual sources of news/info. All stuff that was out there, important, and ripe for the picking — but not yet really expanded on by pubmedia reporters.

Examples of posts about studies:
“Why Many Older People Are Protected Against H1N1”
“A New Way to Create Flu Vaccines”
“Four New H1N1 Studies”

Examples of posts about angles:
“Was H1N1 Info Communicated Well to the Public?”
“Flu.gov: Using Social Media to Boost Outreach”
“The Story Behind Facebook Flu Fighters”
“How Effective Are Flu Vaccines?”

Examples of posts about nuggets of government info:
“National Influenza Vaccine Week: 10-16 Jan”
“H1N1 Timeline from Flu.gov”

Examples of posts trying to illuminate the H1N1 data:
“H1N1 v. Swine Flu: The Rematch”
“Comparing Mortality Data for H1N1 and Seasonal Flu”
“Google Flu Trends Map”

Examples of posts highlighting sources of news and information that reporters might not yet have been aware of:
“Mayo Clinic: A Source of Public-Health Information”
“Assessing H1N1 Blogs”
“H1N1 Video and Audio PSAs in 12 Languages”
“CDC Media Briefing Transcripts”

5. FluPortal news: We used the blog to spread the word about FluPortal news (e.g., about upcoming webinars or new elements of the website).

Examples:
“New Resources and Tools on FluPortal”
“Jan 28 Webinar: FluPortal.org and Crisis Coverage”
“New NPR Flu Map Available from NPR”

6. Photos and video from social media: Flickr, Vimeo, and YouTube occasionally yielded smart user-generated content. (FluPortal’s social-media guide explains how we searched and filtered the massive amount of UGC: http://www.fluportal.org/enrich-your-reporting-with-social-media-leads-and-content/.)

Certain flu blogs were useful in pointing to good news and information about H1N1, but the blogosphere didn’t offer many interesting first-person accounts from flu sufferers.

When we found good UGC, we posted it on the blog — so that journalists could use it to illustrate stories or to generate story or guest ideas.

Examples:
“Photos You Can Use: Swine-Flu Street Art”
“Reporting on Cultural Responses to H1N1”
“H1N1 PSA Videos for Your Website”

In addition to offering up UGC that journalists could use, we wrote several posts explaining how they could go about finding it themselves.

Examples:
“Enrich Your H1N1 Reporting by Using Social Media”
“Find H1N1 Photos for Your Website”

We weren’t able to assess whether pubmedia journalists were using the UGC we found — or whether they were looking for any themselves.

Katie Donnelly, the associate research director of American University’s Center for Social Media, felt that FluPortal should have solicited UGC directly on the site (both crowdsourced information and first-person accounts). Details in this blog post. Something to consider for future “crisis portals.”

7. Generalized resources on crisis reporting: In January, as H1N1 became less of a story, we started thinking about generalizing the lessons of FluPortal for possible future crisis projects. We tried to reflect in blog posts some of that thinking about crisis coverage.

Examples:
“Two Approaches to Integrating Social Media”
“A Guide to Reporting on Crises”
“Crisis Response in Haiti”

Blog Comments
Very few blog readers commented on our posts. We’re not sure why. Site analytics show that people were reading the blog, but for some reason they weren’t commenting.

At first we thought this might be because our blog template didn’t make it clear enough where/how to comment. So partway through the project, we added an obvious link at the bottom of each post saying “comment on this post.” This change didn’t have any effect.

Katie Donnelly, associate research director at American University’s Center for Social Media, told us that in her experience, pubmedia people simply don’t often comment on blogs. She’s not entirely sure why. (Read more here.) We did try offering a Google Group as an alternative place for pubmedia staff to communicate, but only two people signed up.

Bookmarking Using Delicious

We used Delicious.com to collect good articles, audio, and video about H1N1. Delicious allows you to bookmark and tag any web page — and offer up your bookmarks to the public.

The point of the bookmarks — like that of FluPortal generally — was to provide the best available swine-flu information to journalists and web managers working on H1N1. Our collection of bookmarks was assembled with an editorial eye; automated aggregators cannot filter for quality.

We displayed the bookmarks directly on FluPortal’s “Reporter and Program Resources” page and of course in our account on Delicious. Journalists/stations could also subscribe to the RSS feed of the bookmarks.

Throughout the project, we boomarked H1N1 information from mainstream media, public media, public-health agencies and organizations, and even social media.

Once the Daylife widget launched (at the end of January, 2010), it collected pubmedia coverage automatically. We therefore stopped bookmarking pubmedia pieces in Delicious unless the content was especially interesting.

Feedback from stations (solicited by Ken partway through the project) suggested that people weren’t using the Delicious bookmarks as actively as we’d hoped. One person also asked Ken whether the bookmarked pieces were from “trusted sources.” In response, we featured the bookmarks more conspicuously on the “Reporter and Program Resources” page — and tried to explain their utility more clearly. We also changed the way we displayed the bookmarks — to show prominently their reliable source (e.g., NPR, New York Times, etc.).

Unfortunately we don’t have any stats showing how many people looked at/used the Delicious bookmarks (and none, either, showing how many people subscribed to the RSS feed). Future “crisis portal” projects might set up a way to measure this.

Site As a Whole

Goals and Revisions
We revised and edited the site as a whole throughout the life of the project. These were our main goals:

• Make the site look better
• Improve the site’s organization
• Continue adding resources
• Create and add our own guides (to web tools and social media)
• Change the emphasis of the homepage to reflect changes in the H1N1 story
• Improve transparency (about the project and its staff)

This was an ongoing process of tightening/changing language, improving formatting, adding images to break up text, adding links, reorganizing pages, and adding pages.

We worked on this as a team. Feedback from Ken and Rekha’s outreach helped us figure out what needed improving. Two specific examples: We reworked significantly the presentation of information on the “For Station Websites” and “Reporter and Program Resources” pages — to make it more accessible. And we added staff bios on the “About” page.

Josh took care of the more complicated code changes and kept the “Public Media Coverage” page up-to-date (refreshing the features and adding to the list of H1N1 landing pages). He also wrote FluPortal’s technical guides.) I contributed the social-media guide and worked on overall site wording, organization, and consistency.

Some General Practices We Followed
We adopted NPR’s standard of using the terms H1N1 and swine flu interchangeably. (See details of NPR’s decision here.)

We tried to make sure that the links to resources were descriptive — so that people could decide quickly whether it was worth their time to click through. If the name of an organization/article/etc. wasn’t descriptive enough by itself, we wrote up a brief additional explanation.

The links we included on the “Reporter and Program Resources” page came primarily from reliable government, public-health, medical, and media sources. We also added links to some of the “evergreen” posts from our own blog.

The Future

The division of responsibilities within the FluPortal team seemed to work easily and well. Because the four of us were in different places, we used email, IM, and conference calls to communicate.

Four hours a day was just about the right amount of time for the blogger/editor job. If H1N1 had turned into a bigger story, and/or if it had made sense to publish more posts per day, four hours might not have been enough.

A Question About the Blog’s Purpose: Do Reporters Want Story Leads?
If CPB pursues similar projects in the future, answers to the following questions would help steer the blog: Do pubmedia reporters actually want story leads? Are they helpful?

These are important questions because FluPortal’s goal was to help improve reporting on H1N1. Much of the pubmedia swine-flu coverage understandably focused on vaccines, but there were other interesting stories out there that perhaps were underreported. Was this because stations were understaffed? Because they had no specialized health/science reporters? If so, it would be instructive to know how/whether story leads could help.

If it turns out that story leads really are useful to reporters, it would be great if the blogger could do some of his/her own reporting. Not to produce fully reported stories — but to flesh out interesting leads just enough to be more of a tease for pubmedia reporters.

I found the no-reporting restriction to be the most challenging part of the FluPortal blogger job. It seemed to limit the utility and originality of the blog.

Should the Blogger be an Expert on the Crisis Topic?
If at least one person on the FluPortal team had been a health/science reporter, it would have made collecting reputable resources much faster initially. A health reporter would have started with a working knowledge of trustworthy online and offline sources. S/he would likely also have been more skilled at identifying new resources and material for the blog.

For future projects, it would be ideal if one person on the project team — ideally the blogger — were experienced on the relevant beat. I don’t think it’s absolutely necessary, though, as we were able, relatively quickly, to get up to speed on flu.

I do feel that the blogger, even if not a topic expert, does need to be an experienced journalist. It’s easy to spread misinformation on any story; easier still during a crisis because of time pressures. The blogger/editor job requires someone who can distinguish valid sources of information, discern interesting story angles and sources, and write blog posts that may be steering other journalists. A critical editorial eye is key.

Should Future Crisis Projects Use Social Media More Extensively?
We used (and talked about) social media fairly extensively on FluPortal. We didn’t, however, set up Facebook, Twitter, or other similar accounts — largely because FluPortal was aimed at pubmedia, and FB/Twitter seemed too publicly oriented.

Should future projects reconsider this?

One particular social-media tool arrived too late for FluPortal but could be useful to future projects: Delicious now offers a way to send out bookmarks automatically via Twitter. Had this feature existed when FluPortal began, it would have been another useful way to distribute the information we were collecting.

Katie Donnelly — associate research director at American University’s Center for Social Media — feels that future projects could benefit from more crowdsourced information, more first-person accounts, and more ways for station staff to connect with one another. (Details of her critique here.)

What Should Happen to the FluPortal.org Website?
Rather than dismantling the FluPortal.org site to create a more general “CrisisPortal,” why not leave FluPortal.org as is — as an archive — and essentially “copy and paste” it to create a separate CrisisPortal?

Creating a separate generic CrisisPortal wouldn’t be any more work than dismantling FluPortal and re-building it as a CrisisPortal.

Keeping FluPortal online as an archive would serve two goals: it would be a permanent record of CPB’s project; and it would preserve information that could be very useful during a third wave of H1N1 or a future flu pandemic.

Creating a separate CrisisPortal would be easy to do using FluPortal as a model. It could start with (and improve on) FluPortal’s template for site structure and layout. It could even adopt generic versions of many of FluPortal’s tech/social-media guides — which could be useful during any kind of crisis. CrisisPortal could exist either as a permanently generic site, or it could itself be a model to be “copied and pasted” to create other crisis-specific sites (e.g., EarthquakePortal).