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 Technology and Design

by Josh Andrews

Overview

I joined the FluPortal team in the summer of 2009, at the early stages of the project. My initial focus was on the planning and development of the FluPortal.org website as well as curating the For Station Websites section of the site. After the launch of the project, I regularly contributed to the FluPortal blog, maintained and updated the website, and oversaw the online strategy and product management for the project.

Building FluPortal.org

The first phase of planning the website was to determine what content it would house. We gathered all the information, resources, documents, and digital content having to do with H1N1 that we could find. We also reviewed a number of media websites created in the wake of the initial outbreak of swine flu in the Spring of 2009. We created a “scratch pad” using Google Docs where team members could add content and catalog notes and links as we found them. We then went about editing and organizing this content. This initial organization of content informed our decisions on the page structure of the FluPortal website.

With the large amount of material available, the presentation and organization of information was a challenging part of constructing the website. To allow users to easily identify and navigate to the content they were looking for, we defined the main content pages by audience. Web managers would be drawn to For Station Websites and content producers would gravitate to the Reporter and Program Resources page. Throughout the project, the FluPortal team solicited input from members of the public media community. This feedback led to regular improvements in the categorization, organization, and display of the content on the site.

In designing the visual presentation of the FluPortal.org homepage, I used an online service called iPlotz to build a wireframe of the layout. iPlotz offers a basic, free version of the service (which we used) as well as a more advanced set of features available on a cost basis. We planned the layout of the wireframe to serve users with varying needs and topical interests (web managers, reporters, the public). We wanted the homepage to communicate to users a clear explanation of the project’s scope and goals as well as the breadth of content contained on the site. The wireframe for FluPortal.org’s homepage went through 3-4 iterations before the team was satisfied with the plan for the website. Reviewing the wireframes today shows that most of our original planned elements made it to the final product.

The WordPress Platform

The FluPortal website is built on the WordPress publishing platform. WordPress is a well-established open source blogging platform. Widely adopted by individuals and organizations alike, and its large user base provides an ample support community. The WordPress software is free and, for the FluPortal project, was installed on Public Radio Exchange servers by the PRX IT staff.

To save time and resources on web development, we identified and acquired a premium WordPress theme from a reliable vendor. The theme license was purchased from the company Woo Themes at a cost of $125. The theme was chosen for its resemblance to our initial homepage wireframe and for the functionality it offered. After the theme was installed, I went about building the page structure of the website, calibrating the default WordPress settings, and populating the site with project specific content.

A graphic designer was hired to customize the visual elements of the theme as well as to design a project logo. I worked directly with the graphic designer to implement changes to the website’s style sheet.

The php code for the WordPress theme was customized to meet specific project needs, relocate content modules, and make the site more visually appealing. Customizations include:

• Website header — Height of header adjusted; custom logo inserted; email subscription option added to RSS button
• Website footer – Removed redundant page navigation
• Homepage – Removed default page layout below the main slider and added custom code to alter the display of blog excerpts and thumbnail photos.
• Blog – Customized text display settings to display full post content on blog page; changed photo display settings on blog page and individual posts; altered commenting display on blog page and individual posts
• Public Media Coverage page – Created a custom two-column page template to display necessary content (later revised back to a one column layout)
• Two plug-ins were employed on the site for additional functionality — cForms II contact form plug-in; All-in-One SEO for search engine optimization plug-in

While the use of a prebuilt theme did speed up development time overall, we found that the theme did have drawbacks and limitations. For example, we found the layout did little to encourage user comments and downplayed the ones that did come in. We improved what we could with the theme, but adapting any prebuilt template often involves compromises.

I recommend the continued use of WordPress for similar public media crisis response efforts. In addition to its low cost, WordPress can be deployed quickly and provides an easy-to-use interface for non-technical staff. A central managing organization should be identified to support these efforts and oversee web hosting and the use of WordPress MU (multiuser) to deploy new issue specific sites based on a shared or common web back-end.

To combat the trade-offs that come with using prebuilt website templates, a public media WordPress theme should be commissioned and built for future crisis response initiatives. This theme would be designed with the needs of local stations in mind and provide advanced functionality to accommodate multimedia content and social media integration. Having a public media WordPress theme on hand would avoid the process of identifying, licensing, and customizing a premium theme with every instance of crisis response. Many stations don’t have web development staff available and turnkey solutions designed for public media crisis response would make web resources more accessible to the public media system.

Some of this work towards a public media WordPress infrastructure is already in place. In 2009, WNET in New York partnered with Tierra Innovation, a web development group, to build an open source WordPress toolkit. The toolkit includes a number of supported plugins for Wordpress websites that offer features and functionality especially suited to public media outlets.

Such a toolkit should be expanded and combined with a public media website template (and, perhaps, a catalog of visual elements). This has the potential to provide an infrastructure for rapid website deployment options for local stations with minimal web development support.

A similar crisis-website project aimed at local governments has been launched on the Google Sites platform. Stanford University’s Social Innovation and Entrepreneurship Program coordinated the development of emergency response templates available for use on any Google Sites website. In its current stage of development, Google Sites is not powerful enough as a content management system for most public media stations, but this project highlights the potential of turnkey solutions for crisis response efforts.

Resources for Station Websites

The For Station Websites page on FluPortal.org houses a variety of online resources, embeddable digital content, and web tools to help local stations bolster their online coverage of the H1N1 pandemic. Much of this content was identified by the FluPortal team during the initial planning phase of the project. As I curated this section of the website, my goal was to expose local stations to a wide breadth of ready-made content modules as well as innovative approaches and tactics for using free web tools and social media platforms. I kept in mind that many web managers don’t consider themselves programmers or coders. I wanted to provide that audience with quality content that was easy to use and integrate into a station’s website. At the same time I hoped to provide gentle encouragement (along with detailed instructions) on how stations might stretch the limits of their comfort zones in web coverage.

The project’s digital tools included a variety of H1N1-related content, including RSS feeds, FAQs, widgets and badges, maps, and social media feeds. A number of public media stations actively used FluPortal’s digital offerings when building their H1N1 landing pages in anticipation of the second wave of the pandemic. We cataloged these landing pages on our Public Media Coverage page. The simplest contained widget placements; the more comprehensive combined local and national coverage of the pandemic.

FluPortal Guides

To encourage online innovation and experimentation at stations, I included three introductory guides to free web tools at the time of launching FluPortal.org. While these tools were not specific or limited to H1N1 coverage, I hoped that the pandemic would provide station staff an opportunity to try new approaches in building web services. I wrote introductory guides to the following services: the NPR API, Yahoo Pipes, and the Twitter API. Each one explained how a local station could get started using these tools, how that tactic could enhance its H1N1 coverage, and best practices to follow.

The guides were well received and soon became a growing part of our online offerings. We saw the guides as an opportunity to push stations forward and implement new technologies on their websites while also serving FluPortal’s overall goals. As the project continued, we expanded on the existing guides and introduced new services to the roster, including Facebook Connect, CoveritLive, and Delicious, along with a guide to using social media platforms in editorial reporting. (You can see all the FluPortal guides here.)

These guides should serve as a model for a larger clearinghouse on best practices for emerging social media tools and technologies for the public media system. Many public media stations have yet to embrace social media and emerging technologies and lack the resources or time to investigate each new platform. A public media oriented website or wiki could provide stations the needed introduction, information, and encouragement to begin experimenting. This service should encourage collaboration among public media organizations and the establishment of metrics to measure impact from new online tools.

Quick Start Menus

Following the initial launch of the FluPortal website, feedback from stations suggested that the volume of web tools was somewhat overwhelming. The number of options made it difficult for a station to know where to begin. The FluPortal team discussed possible solutions and formulated plans to create three Quick Start menus.

The idea behind the Quick Start menus was to offer tiered combinations of H1N1 web modules to meet the varied needs of public media stations. Depending on the time and technical resources stations had to devote to H1N1 coverage, the menus would provide a clear path to follow to identify the right content for online needs.

The levels were dubbed “Quick,” “Easy,” and “Go Big.” Quick was for organizations that just wanted a widget or two to place on the sidebars of their station websites to provide the latest national H1N1 news for their online audiences. Easy provided tips on bringing more comprehensive ready-made content to a local station website, including H1N1 FAQs, maps, and content modules. A station could populate an H1N1 landing page without devoting any local editorial time or staff to the project. Go Big was aimed at stations that were producing local editorial coverage of swine flu and were looking to highlight that work and integrate it with national news and multimedia content.

The three menu options became the main focus of the For Station Websites page and linked users to detailed instructions for implementation along with examples of how other public media stations had approached their online coverage. The remainder of the For Station Sites page was also reorganized by simplifying and reordering the categorization of the content.

NPR Flu Map

The FluPortal team worked with NPR on the production of a flu map. Launched in November of 2009, the embeddable map displays the prevalence of flu-like illness across the United States by region. The map’s source data comes from official CDC estimates of flu-like illness published weekly. The FluPortal staff worked with NPR’s data visualization team planning the scope of the project and provided feedback to NPR throughout numerous iterations of the final product during its development. When the map launched, it was placed on the FluPortal.org homepage, and we provided marketing support and system outreach on our blog and during our webinars.

Public Media H1N1 Widget

FluPortal played a more direct role in the planning and production of the Public Media H1N1 Widget, again in association with NPR. The JavaScript widget aggregates H1N1 editorial coverage from a large number of sources within the public media system. The widget provides a window into how local communities across the nation are addressing the threat of the swine flu pandemic.

The widget uses Daylife, a content syndication and curation service, to aggregate public media H1N1 content. NPR uses the Daylife service to power the topics.npr.org section of the NPR website. For the Public Media H1N1 Widget, Daylife scrapes content from a preselected list of station feeds and grabs articles with references to H1N1 (and related terms). The widget auto-updates with the latest stories and orders the content by publication date.

We worked closely with the NPR product manager in the planning phases to determine the functionality and design requirements for the widget. The FluPortal team also curated the source list of public media feeds for the widget to scrape. We decided to focus on feeds from approximately 40 public media stations and producers that had shown a commitment to covering the H1N1 pandemic. We limited the number of source feeds to reduce the risk of stories unrelated to swine flu appearing in the widget (while Daylife is largely accurate in its ability to identify stories by search terms, it is an automated service and prone to occasionally mistakes). As the widget neared completion, the FluPortal team oversaw quality assurance and final design changes.

An important note: the Daylife widget is designed to be adapted to future crisis response initiatives. The code can easily be altered to produce a public media aggregation widget on any imaginable topic. In addition, the station source feeds can be expanded or reduced as needed. The only variable about future use of the widget is the continued relationship between Daylife and NPR.

A note on timing: Both widgets suffered from their late arrival to the project. The NPR flu map debuted in late November and the Public Media H1N1 Widget appeared in January of 2010. Development of both widgets began after FluPortal.org site launch, and by the time they were deployed the flu was on the wane and station uptake appears to have been low. It’s possible that had the pandemic continued to worsen, stations might have been more eager for these widgets and their late introduction to the site would not have mattered. Generally, however, we would recommend that major site features should be defined before launch when possible and deployed early on, when interest in the site is greatest. Grant timelines should take such needs into account. While both are high-quality online tools, they were underutilized by stations.

Another challenge with these two projects was a lack of metrics. Neither widget was equipped with any tracking code to indicate if and when it was embedded on station websites. We can only rely on anecdotal evidence about the usage of both of these tools, which is unfortunate given the hard work that went into them.

As the FluPortal grant period comes to a close, we are working with NPR to archive both the flu map and the Daylife widget and document the production of each. FluPortal will reach out to the public media system to communicate that the map and the widget will no longer be supported past the current flu season (the planned end date for support on the widget is May 11, 2010). For the Daylife widget, the javascript code will remain active and, as mentioned before, able to be deployed in other scenarios