Having been lucky enough to have attended the recent Digital Leaders salon “Top Tips to Avoid Failing Badly with Agile Delivery” in London I thought I would share some thoughts on the topic.
One of the things that was discussed was the value of being able to identify what actually is failure in an agile project. Too often, expensive, talented teams of people are spun up to build a service or deliver a product without clear definitions of what success actually is. Without this clear picture of what success is, how can a team know if it is failing? Success criteria are an essential part of agile development, as is the ability to measure against them. If you can’t measure against it, it’s probably not that valuable, and there would have to be questions raised over the validity of it as a success criteria.
Also important is to plan. Planning is often (incorrectly) considered to be something that is neglected in agile delivery. A good product owner will not only have a clear vision of what they want to achieve, but be able to explain that vision. This vision needs so well understood that any member of the team should be able to explain it to anyone. This vision can then be translated into a roadmap, which can be used to plot milestones, measures and challenges that are known. This roadmap is a guide, as any map is, without directions and an understanding of where you are now, a map is just a picture. It is not the turn by turn, step by step guidance to a destination, but it is a very useful tool to establish where you are on a journey. Your product roadmap is never a finished item, this is important to remember. It should constantly evolve based on the things you learn as you build, apply any new learnings, new data or new requirements to it as the become known.
One of the other subjects discussed was the importance of understanding your users. There is a reason this is the first item on the Government Digital Services Design Standards – because they are the most important thing you can do to avoid failure. User research should be the number one thing on your to do list at the start of any project. Key to effective user research is knowing who your users actually are – not who you believe them to be – once you understand the true user base (and you’ll need the data to back that up), then you can start to explore who they are, what they need or use your products and service for and what they think of them currently. Then you can build something, test it with your users and get their feedback. This then powers your next iteration, which is subject of your next research session and this constant feedback loop gives you the knowledge and data you need to be successful.
The most important factor for me though is the ability to measure what you are aiming for. Without metrics you could well be storming along enjoying the most “successful” project of your career without ever knowing if you are actually succeeding. More importantly, where these metrics will be absolutely worth their weight in gold, they will show you when you are failing. This is important. Failing is not the worst thing in the world, as long as you know you are doing it, because then you can stop, change, learn, adapt and turn that failure around. You may even decide to stop completely. Do it fast enough and this is failing well, and it’s by far the best way to avoid failing badly.