Making the Cut

“I put the ticket in more than two years ago and it still isn’t done. When will you get to it?” Our Product lead apparently looked at our backlog as a queue, in which everything that goes in is implemented sooner or later. That makes sense if your tickets are for customer support. But not for a product backlog.

I replied that if her request had failed to achieve priority in two years it was never going to get priority. Other things arrived, maybe before and certainly after, that jumped ahead in line.

Following that encounter I announced to everyone that in two weeks I would purge all tickets older than 18 months. If they wanted to keep an item in the backlog, all they needed to do was to update the ticket. They could simply add a “Keep this” comment.

When the time for the purge came, about a dozen old tickets among a couple thousand had been updated. The rest were purged. And after that I repeated the exercise, bringing the threshold down to 12 months. Nobody missed the purged items.

Backlog, Not Queue #

The idea that a product backlog is a queue in which every input is acted on overlooks the main value of a backlog: things drop off the bottom. A backlog delivers value by determining what work will not be done, as well as by prioritizing the work that makes the cut.

One fundamental problem shared by virtually all electronic product backlog tools is that they remember too much. No doubt you have had that conversation in which someone says “Archive it in case we need to reinstate it later.” or “Let’s save it so that we can look back at how we reached that decision.” It sounds good. But the standard critique of waterfall planning applies here. When you reinstate or look back at an item you have a different, more fully informed context. Your earlier considerations lacks the benefit of the learning you did since setting aside the item. Picking up that old work again requires mapping it into your current context, and working your way through those same considerations in light of your current knowledge. Reviving your reasoning or the ticket itself is not so straightforward as it first appears. A good rule of thumb is “If it is important enough, it will come up again.”

Items drop off a backlog not because they embody bad ideas, but because they are not good enough. An individual item cannot be measured in stand-alone fashion on its own merits. Instead it must make the cut in competition with the other items in the backlog. The decisions about what to prioritize and where to draw the line are not easy. Nevertheless, they are key to achieving focus on the items that will be developed.

Trade-Offs #

When your backlog is large, you spend a lot of time and effort on reviewing, updating, and reprioritizing it to reflect new learning. And you will invariably change or set aside some of the items in the backlog after this investment of time and effort.

Those changes and set-asides are not necessarily dysfunctional. Completing every item in your backlog without changes suggests that you are not learning as you go, because learning modifies and improves the way you approach work. And learning that an item is not worth doing frees you to do more important work. So changes and set-asides can be valuable. They can also be taken too far. A high rate of set-asides suggests that your backlog is either too large or too ill-defined. The sweet spot sits in the middle.

How big and how well-defined should your backlog be? This is a question of trade-offs. Do you spend inordinate amounts of time on review, updates, and prioritization? That suggests that your backlog is too large. Do you regularly radically change backlog items? That suggest that your backlog is ill-defined. Do you set aside a high proportion of your backlog items? That suggests it is too large and also ill-defined.

One rule of thumb is to size the backlog by the look-ahead it provides. Make it large enough to include work you expect to complete in some reasonable amount of time plus work coming up in the next such interval.

Conclusion #

The backlog is your tool for managing these trade-offs. Its main benefit is giving you a place to make the cut, and set aside the things that don’t belong in view now. Make the cut, give yourself focus, and don’t fall prey to the temptation to remember everything and your flow will improve.