A Publishing System That Still Works on Your Worst Days

Set your publishing up so a bad day asks you for no decisions. Write a week of posts in one sitting, keep at least a week queued ahead, and keep a short folder of five-minute posts for the days the queue runs dry.

Most posting routines are built on a good day. You wake up rested, you've got an idea, you have 30 quiet minutes. It works for two weeks. Then a deadline eats a Wednesday, you catch a cold, and by Friday you haven't posted in four days and catching up feels pointless.

That's not a discipline problem. The routine assumed conditions you don't always get. This post is about setting things up so you don't need them.

Goals versus systems

"Post three times a week" is a goal. It tells you what to hit and nothing about how. James Clear's version of the distinction is the one most people quote: "Goals are good for setting a direction, but systems are best for making progress."

A system for posting answers four questions ahead of time:

1. Where do ideas go when you have them?
2. When do you write, and for how long?
3. When do posts go out, and to which networks?
4. When do you look back at what worked?

If all four have answers, a bad week mostly runs on work you've already done.

Count the decisions in your current routine

Write down every step between "I should post" and a post going live. For most people it's something like:

1. Open a blank document
2. Decide what to write about
3. Write the post
4. Edit it
5. Open the network
6. Paste it
7. Find or make an image
8. Write alt text
9. Pick a time
10. Publish
11. Do it again for every other network

Steps 2, 4, 7 and 9 are decisions, and step 2 is the one that eats a bad morning. For each step, ask in order: can I drop it, can a tool do it, can I do it once a week instead of every time?

• Drop it: you probably don't need a custom image on every weekday post.
• Hand it to a tool: a scheduler covers opening each network, pasting, picking a time and publishing.
• Batch it: choosing topics, writing and editing can all happen in one weekly session.

What's left for a bad day is short: open your drafts, pick one, schedule it.

A note on the "willpower budget" you'll see in a lot of productivity writing, the idea that every decision drains a fixed daily supply. It's a popular explanation, but the evidence behind it is shaky: a preregistered replication run in 23 labs with over 2,000 people found an effect close to zero. You don't need the theory. Fewer decisions on a bad day is simply less to get stuck on.

Write once a week

Block 90 minutes once a week. In it:

• Pick five topics from your idea list
• Draft all five, one after another, without editing
• Edit all five in a second pass

Keeping drafting and editing apart is most of the trick. Drafting is about getting raw material down; editing is about cutting it. Doing both at once slows both.

Keep an idea list

The session falls apart if you sit down with nothing to write about. Keep a running list and add to it whenever something comes up:

• Questions people ask you, in replies, DMs or calls. They're the safest topics you have, because someone already wanted the answer.
• Things you noticed in your field this week, and what you think about them.
• Things you did: a mistake, a process you changed, a number that moved.

It helps to group ideas under three or four topics you keep coming back to. A freelance designer might use design process, pricing and clients, tools, and career lessons. Groups make the list easier to scan and stop you writing about the same thing five weeks running.

Keep more ideas than one session needs. When a week goes badly and you add nothing, the list still has enough for the next session.

Queue a week ahead

Once a post is scheduled, it goes out whether you're at your desk or in bed with the flu. That's the single change that does the most for consistency.

After the writing session, spend 15 minutes scheduling the week. Adjust each version for its network while you're at it: X and Bluesky need the tightest copy, LinkedIn can take more structure, and Mastodon readers expect alt text on every image and a content warning on some topics.

Then keep a buffer: always at least one week of posts scheduled beyond today. If a week goes badly and you can't write, the queue covers it, and you've got seven days to recover before anyone sees a gap. Two weeks covers a holiday.

If you're worried that scheduled posts reach fewer people, the tests we found say they don't. We went through them in does scheduling posts hurt reach?

Keep a folder of emergency posts

Some days the queue is empty and you can't write. Decide now what you'll post on those days, and write those posts now, because your bad-day brain won't.

Keep five to ten short posts in one note. Some shapes that work:

• A question to your followers about something they're working on this week
• "The most useful thing I learned this month," filled in from your notes
• A customer message, shared with their permission, and one line on what it taught you
• An older post that did well, quoted with one sentence on what you'd add now

Each takes a minute to finish and schedule. They're not your best work and don't need to be. They keep a bad day from turning into a silent month.

Look back every other week

Keep the review small. A note with three lines:

• What did best, and why do you think it did?
• What was easiest to write? Can you do more like it?
• One change for the next two weeks: a different opening, a new time, a format you haven't tried.

Then check the system itself, not just the posts:

• Is the idea list growing faster than you use it? If it keeps running low, you need more inputs: more conversations, more notes as you work.
• Did the queue ever run empty? If it did, the writing session is too small or too easy to skip.
• How many networks does each idea reach? If most ideas become one post on one network, you're doing the hard part and skipping the cheap one.

Counting posts tells you the machine is switched on. These tell you whether it works.

When things go badly anyway

When you can't do the writing session:

1. Check the queue. If the rest of the week is scheduled, do nothing.
2. If it's empty, pick one emergency post and schedule it.
3. Put the writing session on tomorrow's calendar. If tomorrow's bad too, repeat step two.

After a rough stretch, like an illness or a trip, don't try to catch up. Fill any gaps in the queue, book the next writing session, and carry on from there. Nobody is keeping a tally of the posts you missed.

The whole thing on one page

In Shaflex you set your weekly time slots once, then pick a day and a slot for each post, and the queue publishes on its own in your timezone. The Composer gives each network its own copy of the draft with its own limit, which covers the per-network edits. To get more reach out of each idea this system produces, see the content flywheel.

Sources

• James Clear: Forget About Setting Goals. Focus on This Instead.
• Hagger, Chatzisarantis et al. (2016): A Multilab Preregistered Replication of the Ego-Depletion Effect, Perspectives on Psychological Science 11(4)