IIWO / ROUTINE-TASK ECONOMICS

Is it worth optimizing?

How long can you work on making a routine task more efficient before you're spending more time than you save?

Based on an xkcd comic

01 Inputs

Time saved

Frequency

times a

Time period

Hourly rate

$
/hour
02 Break-even budget

Time saved / year

5.1 hours

Which means you could use that much time to optimize the task before you would be spending more time than it's worth

Monetary value

$101.39

Which means you could spend that amount to optimize that task

03 / CONSTRAINT

Adjusted for time scarcity

Your time is more valuable when you work a lot and have little free time. The value below multiplies the base monetary value by the ratio of work hours to free hours.

Hours of available free time a day

hours/day

Hours of working a day

hours/day
04 Scarcity-adjusted output
$405.56
Adjusted monetary value (base × 4.0 work/free ratio)
0.7%
This optimization is worth 0.7% of your free time per day

05 / PLAYBOOK

Ideas for optimization

Strategies to consider when optimizing routine tasks.

Keyboard shortcuts

There's keyboard shortcuts that anyone could benefit from. You don't have to be a programmer or a video editor to save time with shortcuts. Even something as simple as Ctrl + Backspace to delete a whole word at once will add up over time.

Snippets

Using snippets when coding can save a lot of time. I have a snippet for every import I commonly use and also for things like creating a skeleton of a React component.

Text expansion

This is basically snippets outside your editor. A short trigger can write your email address, a common reply, an issue template, or a command you use all the time. It is worth setting up when you keep typing the same thing in different applications.

Templates and checklists

If you make the same kind of document, project, pull request, or meeting note over and over, start with the structure already in place. A template saves typing, but it also saves the time spent deciding what to put where.

Extensions

Of course this depends heavily on which software you work in. There's Emmet, which is a well-known extension, but my favorite is Jumpy, which I can use to quickly jump to a different place in the text. Vim has a lot of "motions" that you can use for the purpose of jumping around quickly in the text but Jumpy provides a single quick shortcut to achieve 90% of the motions you could want.

Batching

Doing similar tasks together means you only have to get into the right application and state once. It works well for things like processing receipts, resizing images, replying to email, or reviewing a stack of pull requests.

Better defaults

The normal action should give you the result you usually want. Set your editor, browser, and operating system up so the right folder, export format, naming scheme, or window layout is already selected instead of being corrected every time.

Remove a step

Before automating a task, check whether the task needs to exist. You might be copying data between tools that can already integrate, generating a report nobody reads, or fixing an error that a better source setting would prevent.

Macros and scripts

A macro is useful when the clicks are always the same. A script is useful when the files or data change but the operations stay the same. The best candidates are boring tasks you can describe as a list of exact steps.

Reduce rework

A saved second is nice, but avoiding a recurring mistake is usually better. Clear names, validation, presets, and a small checklist can stop the same correction from coming back every week.

Faster navigation

A launcher, bookmarks, file search, and window shortcuts save the little pauses between tasks. I use them to get to the next file or application directly instead of hunting through folders, tabs, and menus.

Precompute and cache

If the same expensive work happens every time, do it once and reuse the result until its inputs change. That can mean keeping a build cache, saving a search, exporting a reusable asset, or leaving a frequently used project ready to go.

More powerful computer

Yes, even something as radical as buying a new computer makes sense. The math checks out. You have to load something literally hundreds of times a day and savings of a couple seconds each time adds up quickly.

An average computer should last you at least 3 years. Over that period, 1 second savings 100 times a day at $30 an hour gives you $900 to spend. Can you get a better computer that achieves that saving for that amount of money? I know that I could but if you're already on the latest generation you might not.