The why and how of daily updates

In August 2018, my work colleague Tomo Kazahaya and I came up with the idea of using GitHub issues to keep track of daily work checklists. We created a private GitHub repository called daily-updates. In October, I copied over the idea to managing personal checklists; you can see my first issue created October 20 or my issue for yesterday.

Since then, I’ve been creating daily updates issues on my personal repository on any days where I’m doing meaningful work on personal projects or tasks beyond daily personal maintenance.

In this post, I’ll talk briefly about the following:

  • The benefits of a broadly visible written document for a personal checklist

  • The benefits of daily granularity for a personal checklist

  • The benefits of using GitHub issues to record daily updates

  • Some protocol questions I don’t have clear answers to

The benefits of a broadly visible written document for a personal checklist

I’m using the term “broadly visible” instead of public, because in some cases (such as companies) daily updates of employees refer to a lot of confidential company information, and therefore can only be visible within the company. In this context, “broadly visible” would mean visible within the company.

The emphasis on “broadly visible written document” is to contrast with two other kinds of approaches:

  • Private personal checklists, that each person manages on their own, but are not visible to others: The advantage of a broadly visible document rather than a private personal checklist is that people can check any time what the others are working on. This allows them to offer thoughts and feedback, and also gives a better organization-wide picture of what is going on.

  • Daily stand-ups, where everybody describes what they plan to do during the day: Written documents are preferable to daily stand-ups because they can be continuously updated as needed, can be referenced any time, and work better for teams that don’t all start work at the same time. Also, my historical experience has been that most people hate stand-ups; the Internet seems to agree.

A dilemma that sometimes comes up is that some subset of one’s daily tasks needs to be kept private, and therefore cannot be explicitly included in a broadly visible daily checklist. I’ve encountered this sort of situation both in my work checklists and non-work checklists.

For these kinds of situations, I’ve generally found it good enough to include stand-in checklist items that may not accurately describe exactly what the underlying task is, but frame it in a way that, when I read it, I know what it is. For instance, a confidential meeting may be written up as “meeting to brainstorm ideas” or something similar, which is vaguely true and good enough to keep track of things.

If most of the items in the checklist need to be highly confidential, though, broadly visible checklists may not make sense. It may therefore be better to restrict the visibility of the checklist to just yourself and one or two other people who collaborate closely with you.

The benefits of daily granularity

Checklists can be made at different granularies: hourly, daily, weekly, monthly. I have found that daily is a good unit for a certain kind of checklist.

In my experience, a day is long enough and flexible enough to have a reasonable set of things to fit in (without being too prescriptive about what to do at what time) while still short enough to plan even in chaotic environments.

For larger teams and/​or more complex projects, planning needs to be done on longer timescales, but a flat checklist of items may be too simplistic for longer timescales, and tailored project management tools (such as Jira) may be better suited to such planning. Such planning can be complementary to the daily checklist.

Why have a daily checklist if we have broader project planning? A few reasons:

  • Daily checklists can include items that are not part of the broader plan, but more one-off pieces of work.

  • Daily checklists offer a clearer sense of how the work breaks down in practice at a daily granularity. Project management tools can be quite complex, and while they may be better at overall task tracking, they may not be that good at getting a sense of “how much work of what kind can be done within a day?”

The benefits of using GitHub issues to record daily updates

Using GitHub issues for daily updates is a bit of a hack, since that’s not the original purpose of GitHub issues. Nonetheless, the hack has worked well in the two contexts where I’ve tried it (work-related checklists at my company, and personal checklists). Here are some of the advantages of GitHub issues for daily updates:

  • GitHub has good checklist/​checkbox support, which makes it easy to track progress. On the main issues page, one can see at a glance the number of checklist items per day and how many of them were completed.

  • The concept of closing an issue helps create a sense of finality: once somebody has finished checking boxes and updating their daily updates issues, they can close the issue. Then others know that the issue represents the final state of what was done during the day.

  • If much of the other work being done is also recorded in GitHub, GitHub’s support for linking within GitHub can make it easy and fast to add links to the actual work. This could include links to GitHub issues, commits, or pull requests.

  • It’s easy to see daily checklists for many different people together.

Some protocol questions I don’t have clear answers to

Some of the questions I’ve wondered about, but don’t have clear answers to, are below. There are pros and cons of various approaches and I’m guessing the optimal answer varies by context.

How should we handle stuff that came in after the original checklist was made?

My approach is to just add it to the list.

If it’s particularly important to know what fraction of work is pre-planned versus spontaneous, the stuff added later can be put in a separate section for spontaneous tasks. I’ve tried this in the past, back when getting a sense of the level of pre-planned versus spontaneous in the task mix was important to me. However, I nowadays organize my checklist by topic instead.

Should boxes be checked if the whole task as intended wasn’t completed, but part of it was?

My current strategy is to check the box and edit the description to note that only part of the work was done.

Should people be expected to keep the checklist updated throughout the day, or only once when closing the issue?

I think that as an organizational requirement, there should be only a requirement to create the checklist (open the issue) early in the day, and close it either at end of day or early in the next day. People who want to update the checklist more frequently are welcome to, but there is no requirement. For some organizations, there may be good reasons to impose a requirement for more real-time up-to-dateness.