How Do I get through a 500 item backlog? One at a time
Dec 28, 2021
TL;DR
It’s straight-forward, but not easy, to get through Bugs
Your progression through a long backlog will grow faster geometrically as you start to recognize patterns and trends. The trick is to expose yourself to the backlog frequently enough and deeply enough, otherwise you’ll forget them
For a backlog of 500, dedicate 1 hour per day for 1 week on it. See how far you get in the first day, then the second, and make a judgment call on whether you’re getting faster. If you are, just get through to the end of the week
Know that you will face many “ugh-fields” along the way, so don’t be surprised or discouraged. Treat them as you would enemies in a video game: a signal that you’re making progress
More about ugh fields, and tactics
I have written some handy Canned Phrases/Auto-responses to help you with the task
Here’s a great post about Embracing the Grind.
My favorite part is toward the end, where the author describes how they got through 2000 open bug reports. Up until them, nobody could do it—or to be more accurate, nobody had the appetite to do it.
I printed out all the issues - one page of paper for each issue. I read each page. I took over a huge room and started making piles on the floor. I wrote tags on sticky notes and stuck them to piles. I shuffled pages from one stack to another. I wrote ticket numbers on whiteboards in long columns; I imagined I was Ben Affleck in The Accountant. I spent almost three weeks in that room, and emerged with every bug report reviewed, tagged, categorized, and prioritized.
This is a great story. It reminds me of [Designer], and how one day, there were just… big foam boards all around the office. Sam said, “I had trouble visualizing the flow and logic of what we were talking about [we were discussing the Product Principle of whether or not Explorations should spin off their own Dataset or not], so I just printed them out.”
I don’t know where she got the foam board. I don’t know how late she was at the office pinning all of this stuff together. But she did it.
At the same company, my manager asked me, “what’s so hard about just sitting in a room and doing it? [it being getting through the backlog of Bugs]”
Indeed, what is so hard about it? Some reasons that might come up:
Can’t find time in your calendar—this is a made up reason. Most of those meetings probably don’t matter as much as you think they do. And if you put a 4 hour block on your calendar saying “DO NOT DISTURB—I am getting through the Bug backlog”, my bet is that most people would respect that block. And then when you finish, take the rest of the day off. You’ve earned it
Can’t answer every ticket. Awesome, you don’t have to! Your job is to nurture the tickets to resolution by the people best suited to do so. When you think about it, there are really only three buckets:
You understand the bug, and you know who should own the fix. In this case, assign it to that group. If you don’t know who should own the fix, take your best guess and ask them to help you to reassign to the right person if you’ve gotten it wrong
You don’t understand the bug. In this case, enlist a SME, a mentor, or your manager. It is equally all of their jobs to help you do this task. But do them a favor and have a handful of these ready to go in one short 20-25 minute meeting. Protip: make sure you dig into their thought process and reasoning so that you can lessen the times that this happens
Not enough information to make a determination—when you have this case, ask clarifying questions to the bug filer, and assign it back to them. Ping them a week later. If they haven’t replied, close the ticket saying “I’m going to assume this has resolved or it’s not as urgent anymore, but please reopen if I’ve gotten that wrong”
Things that might fall into this category:
Cannot Reproduce the Bug/no repro steps included in the ticket. If this happens, how is an Engineer going to know whether they’ve fixed it or not? Repro steps should be table stakes for reporting a bug
It is unclear what the expected resolution is. Is it that an immediate fix for the User experiencing the bug is pushed? Or is it that the Eng team investigate because this is something that might be widespread?
Here is a handy dandy Canned Phrase that you can use for dealing with Bug tickets:
Closing as won't do
Decision:
The product team synced on this and to the best of our understanding, we believe that this should be closed as a “won’t do”
Reasoning:
[Put reasoning here in plain English]
e.g. We haven’t seen enough cases of this warrant a fix. It sounds like there are [insert X instances per week] of this
e.g. It doesn’t seem like it’s THAT painful for [User Type]. Our understanding is that they have a workaround for this
As always, we are fallible. We made this decision with our best understanding of the information at hand and are closing this out to give you a firm decision to push back on. If there is information that we don’t have or we seem to have ignored that you feel strongly should lead to a different conclusion, please comment on here and let us know