Delivery

  • A Happening An impermanent experience can be more transformative than a permanent artifact
  • All Citizens Very few problems are shared by all citizens. Take the time to define precisely which groups have the problem you're tackling.
  • As Simple As Possible Processes and messages should be simplified as much as possible, but no further.
  • Bridging tech capacity Is your project doing something that the people who should make it cannot do for time, money, skill reasons?
  • Content Before Engineering Can you explain a process so that people really understand it well before you build the system to contain the process?
  • Correlation Isn't Causation Correlation doesn't equal causation, so make sure you have further research so you don't lead people astray.
  • Data Spills Kill The Environment Avoid projects that simply publish large dumps of data. They don't engage end users and the subsequent failure can chase away partners.
  • Define Failure Define what failure looks like at the start of a project.
  • Digitize a Process Build a tool that makes an existing process easier, faster, better, stronger.
  • Do Your Homework Research your idea first, check if it has been attempted before (successful or not) and ask for advice from those who've already tried.
  • Don't Reinvent The Wheel Build as much of your site as possible with off-the-shelf tools.
  • Don't Wait, Scrape Don't hang around waiting for governments to release open data if you can get what you need some other way (e.g. screen scraping).
  • Eat Your Own Dog Food but don't only eat dog food
  • Eliminate Your Job Automation might replace people in your organization. Don't work around that, have them do something better instead.
  • Faster Horses People, when asked what they want out of a project, will optimize their existing workflow instead of inventing a new one.
  • Fuck It Ship It Publish your prototype as soon as possible, instead of trying to build a full product.
  • Gamification "Badge acquisition" isn't actually "fun” despite being a “game”, it's just addictive.
  • Keep The World Informed Share your experiences with the world as you're building. Help the community learn from you so that you can learn from them.
  • Kill Switch Determine under which conditions you will terminate your project.
  • Launch Without A Home Page Build the meat of your site first, and add the front page last.
  • Launching Is Just The Start When you finally launch, you've only just started. Launching your project is not a destination and it's not finished, you've only just begun.
  • Make it Mobile Take information in a closed format and make it available to people when and where they need it. Today, that often means on a smartphone.
  • Measure The Right Things Gauge your success by counting a metric. Website visitors is not a good metric.
  • Open Your Admin Expose your internal administrator interfaces as much as possible to the public
  • Say No When negotiating with funders, say no to projects which would not work in your local environment.
  • Scratch Your Own Itch The best way of solving a real problem is to address one that you actually have.
  • Start With The Citizen Put the user first and consider the whole service.
  • Think Backwards Start by describing what you want to achieve for your users, and then work out the steps to get there.
  • Worry About Abuse Later Don't build systems to cope with abuse until you actively experience it.