If you are running a method whereby you estimate or size stories on your backlog and then plan chunks of work in time-boxed product increments, what do you do when you need to investigate some feature or technology to satisfy a future piece of work?
You could size the spike just like any other piece of work. You could allocate a size to it like a time-box. You could avoid sizing it altogether.
All these options are valid as long as you still plan the work. You can try all three in different iterations and work out what works best for your team. You should know before you start…
- Why you are doing the spike
- When the spike is “done”
- How much you want to invest in the spike
When the work is in progress, the spike should be discussed and tracked just like any other work. Spikes are likely victims of blocking and can also become rabbit-holes into which time pours endlessly. Including spikes in your daily stand-up means they get the same respect and support as any other task.