At the recent NTEN conference in Washington, DC, Jenn Olivia and I had the opportunity to present with two of our favorite clients - Heidi Blobaum, Product Officer at the International Rescue Committee, and Megan Farrington, Deputy Director of Internet Communications at the Drug Policy Alliance.
Our presentation focused on one of the least favorite parts of managing a website - what to do when everything falls apart. Our official title was "Welcome to the Website Emergency Room: Find and Pinpoint Problems When Everything Falls Apart."
We discussed how to triage problems in the moment, what support tools we find helpful, and what systems and processes you can put in place to prevent frequent website fires. Heidi and Megan offered real-life examples of how they troubleshoot problems during crisis times like end-of-year giving, as well as during new product launches.
For those of you who weren't able to attend the conference, here's the Cliff Notes version of our presentation.
When troubleshooting issues on a site, it's helpful to first identify who makes up your team and provide each person with a clear role. These are the roles we discussed:
- Squeaky Wheel: The non-techy person who finds lots of problems on your site.
- The Technology Whisperer: The “webmaster” at your organization, person who knows the most about how to use the website but isn’t a full-time IT person.
- The Fixer: The person who develops your site and fixes any problems you encounter.
We have a saying at Jackson River that “No one is on the table!” As website managers, we are not surgeons with a human’s life in our hands. So, while we take our jobs very seriously, it is crucial to keep perspective, which can really help with the first rule of website triage: stay calm and breathe!
Key Rules of Triage:
- Bedside manners matter: Whatever your role, a little kindness and patience goes a long way!
- Which one?: Most issues can be assigned to one of three buckets: A bug, a service outage, or user error.
- Search for clues: Having clue gathering tricks in your back pocket and making sure your team has the same tools will make triage more efficient.
- Recreate and document: It is crucial to assure that as your team moves through the triage process, someone can recreate the issue at hand and provide detailed steps for how it was replicated.
When something is going wrong on your site, you want to have a solid system in place for tracking problems and keeping everyone in the loop. We recommend having a clear system of notification when something goes wrong, a path to escalation once a problem occurs, and defined roles and responsibilities for handling the escalated issue. Long term, all sites should plan and budget for regular security and maintenance updates, and should make sure that the support systems in place work for everyone.
Systems and Processes
When there are no systems in place in an organization, everything can become an emergency quickly. We recommend developing regular routines and checklists that your team can implement daily, weekly, monthly, and when making changes to your site.
- Daily: make sure that you have ping alerts set up, as well as server alerts.
- Weekly: check for security updates, website analytics, and webmaster tools.
- Monthly and with site changes: test against the most popular functionality of your site by visitors and test against what your administrators do every day.