When you go with “No Estimation”, what becomes more important is the way you split the work and the way you prioritize it.
Vasco Duarte provides some strategies to prioritize tasks(stories, epics) :
- “Do the highest technical risk first”, commonly used in software,
- “Do the easiest work first”, synonym to quick wins,
- “Do the most important strategic work first”, more business oriented,
- “Do the legal requirements first”, it’s a real pain if you forget about it in some occasion, so doing it first can really help.
- “Liability driven investment model”, which is always something you have to take into account and particularly when the structure is small.
My opinion is that you need to use a combination of all of this. Yes prioritization is hard but you removed some of your work by not estimating so that you can focus on more valuable tasks like prioritization.
Sometimes, what helps prioritizing is cutting the tasks(stories, epics) into smaller pieces. It clearly helps to deliver continuously. That’s also a difficult exercise but you can follow these 6 proposals :
- Split into operations or actions
- Split based on business rules
- Simplify business process
- Reduce options or complexity
- Focus on major effort first
- Deliver basic user interactions first
In any cases, you shouldn’t slice the work with tasks about frontend, backend, databases, etc. One task should cover all these aspects so that when something is done, it’s really done from a user perspective. For more details about these slicing techniques: http://techblog.holidaycheck.com/brilliant-ways-to-slice-user-stories/#more-122
So now, you have some techniques to slice and to prioritize tasks but there is one evident thing which is often forgotten : you should prioritize accelerating-returns-tasks first.
What is an accelerating-return-tasks?
It’s a task which will help you to deliver anything that you do next faster. It’s often a waste, meaning that it doesn’t bring direct value to the customer. It’s often forgotten because it’s often on the verge of the specialities of the team. These tasks are often highlighted during the retrospective or at any point in time if the team is really constant-improvement-oriented. One example of these tasks is to improve the tooling around the delivery of your software so that it becomes a no-brainer-very-fast-low-risk process to deliver anything in production. You can have a look at the previous posts : ship early, ship often.
One other example of accelerating-returns-tasks is moving to the cloud. If you are a Software as a Service vendor, it’s not something which directly benefits to your customers but it’s something that accelerates your ability to develop new features, scale them and control your costs till a certain point. The automation of the infrastructure and now more and more of the platform is clearly something which accelerated and still accelerates the IT industry but only a few persons perceived the interest of this at the very beginning.