Okay, no one needs more meetings but before you stop reading I’m going to make a case for doing two backlog grooming sessions every sprint. I believe this will make your backlog grooming more efficient.
Backlog discovery
Backlog discovery is the first of the two meetings I’m advocating for. This should be SHORT: like 30 minutes or less. Backlog discovery is meant to support backlog grooming. In this meeting you do two things:
- Uncover the uncertainty that the team has about any given ticket
- Assign someone to reduce that uncertainty.
Uncertainty can be anything from unclear requirements to not knowing what lurks in the code in that area. The key here is to figure out what your questions are and send a “scout” ahead to figure it out. That scout should then record their findings on the ticket.
Backlog grooming and estimation
Once the uncertainty of one your upcoming tickets have been reduced, you can get to that actual backlog grooming and estimating. In this meeting, the scout will tell the team what they discovered so that the team can accurately estimate the ticket. This meeting should again be pretty short. Ours are about 45 minutes long. Having two backlog grooming sessions shouldn’t add more total meeting time to your calendar.
Spacing
In order for the double backlog grooming sessions to work, you have to space them apart an appropriate amount. If you can have your backlog discovery meeting early in your sprint and your backlog grooming and estimation meeting later in the sprint, it gives the scout time to actually reduce uncertainty.
I’ve found over time and a couple of scrum teams, that this method works pretty well. You’ll know it’s working if the sprint work seems to run more smoothly and your estimates are more accurate. If you try this method, try to avoid turning both meetings into estimation meetings. It’s an easy pitfall to make! The point really is to reduce uncertainty before estimation to make everything run more smoothly, not to create more work for yourselves.
Kate Moore Hermes 