Editors note: Today’s guest blogger is Ivan Chou, Web Applications Engineer for the American Red Cross. When the American Red Cross was struggling with sub-optimal search functionality, it turned to the Google Search Appliance (GSA) to improve information “findability,” both for internal and external users.
Since its founding in 1881 by visionary relief coordinator Clara Barton, the American Red Cross has been the nation's premier non-profit emergency response organization. Today, in addition to domestic disaster relief, the American Red Cross offers compassionate services, from help for the needy to educational programs.
As one of the world’s largest non-profits, we rely on our public-facing websites and our intranet to keep millions of people and tens of thousands of employees and volunteers informed about our activities. On any “normal” day, we get up to 200,000 hits on our main website – but that number swells to millions during any disaster. The main external website, www.redcross.org, receives significant traffic from people seeking everything from CPR class schedules and Blood Drive locations to information about disasters, such as the 2011 U.S. tornadoes.
With such a heavy reliance on the web for delivering information, the American Red Cross needed a faster search system that would deliver better results. We had been using a solution that came bundled with our content management system (CMS), which we implemented in 2009. That search system used a meaning-based context model, which means results were driven by questions, phrases or sentences rather than keywords – but this approach often failed to deliver relevant results. People had to know in advance, for example, if they needed to search about community services, educational programs, international relief, and so on – it was a lengthy, and not very successful process.
Slow performance and lack of relevant results prompted us to investigate search alternatives – an initiative that coincided with a redesign for www.redcross.org and creation of www.measlesinitative.org, a new site supporting a multi-agency push to help halt the spread of measles worldwide.
Working with one of Google’s solution partners, Fig Leaf Software, we began evaluating our options and calculating the costs and benefits of deploying a new search solution. We were at a tipping point – we would have to pay more in licenses for our CMS to support our sites, but we saw in evaluating different options with Fig Leaf that we could save IT costs and achieve better results by setting up two Google Search Appliance (GSA) systems rather than staying with siloed CMS-based search systems. After substantial evaluations, our outsourced data center deployed two GSAs: one for production and the other as a backup – we wanted a redundant solution so that people could reliably find information in the event of emergencies.
The Google Search Appliance systems were implemented over a single weekend, and they now power search across our employee intranet as well as on the public redcross.org and measlesinitiative.org sites.
Right away, the search results from the GSA were excellent. We did almost no tweaking on our end, and our internal and external users comment on how pleased they are that the right search results come straight to the top, whether the query is about CPR class schedules or disaster relief. Visitors to the intranet as well as to the two public websites now have ready access to information through a powerful, intuitive and familiar search experience.
When we moved to the Google Search Appliance, search just started working, and working very well. We predicted that we could save IT costs and achieve better results by setting up the Google Search Appliance, and that’s exactly how it turned out.
No comments:
Post a Comment