Workflow 25: How to Prioritize Your Day With AI

This is the first thing Albus does at 8am, before he writes the morning brief in Workflow 1.

Running three businesses means everything is urgent and nothing is obviously first.

I'm also easily distracted by the next shiny thing, which I'd love to claim is a strength.

So there's a step that happens before my morning brief goes out. The agent reads everything my other agents filed yesterday, weighs it against the goal I set for the quarter, and proposes what I should do first. Then the brief arrives with that already decided.

What this replaces

Opening your task list and picking whatever feels most pressing.

Which is usually the newest thing, or the thing someone chased you about, or the thing you find most interesting that morning. None of those are the same as the most important thing.

Across three businesses the failure gets worse, because each one generates its own urgent-feeling work. Without something weighing them against each other, whichever business shouted last wins the day.

Blueprint

Everything the other agents filed, measured against the goal.

Trigger: scheduled, every morning at 8am, as the first step of the morning brief. Ella has already cleaned the inbox at 7am. Inputs: my to-dos, my inbox, what came out of yesterday's meetings, and my current goal. Output: a proposal for what to do first, folded into the morning brief.

What it reads:

• Tasks pulled out of yesterday's meeting transcripts
• The relationship records, so anything I owe a person is visible
• My inbox, already cleaned by the email agent
• Whatever I've got in my to-do list
• The goal I've set for this month or quarter

Then it proposes the highest-value thing to do before anything else.

Organize

What has to exist first:

• A written goal for the quarter or the month, somewhere the agent can read
• One to-do list
• The agents that feed it, or at minimum a habit of capturing meeting commitments
• Enough context about your business that the agent can tell important from noisy

That first one is the piece people skip, and it's the entire mechanism. Without a goal written down, prioritization is just sorting by deadline, which is how you end up busy and nowhere.

Mine changes more often than a quarter. Depending on what's happening, it can shift almost weekly. So it's written somewhere easy to update, and the agent reads it fresh every run rather than remembering an old version.

Systematize

You are [AGENT NAME]. Before you write my morning brief, work out what
matters most today.

Read:
- My current goal: [WHERE IT LIVES]. Read it fresh each run, it changes.
- My to-do list: [WHERE]
- Tasks captured from yesterday's meetings: [WHERE]
- My inbox, after it's been sorted: [WHERE]
- My relationship records, for anything I owe someone: [WHERE]

Then build a proposal:

1. THE HIGHEST THING
   One thing to do before anything else today, and why it's that one.
   Tie it explicitly to my goal. If it doesn't connect to the goal, say
   so and tell me why it still wins.

2. WHAT ELSE MOVES THE GOAL
   Two or three other things that actually push the goal forward.

3. WHAT LOOKS URGENT BUT ISN'T
   Anything loud that doesn't move the goal. Name it, so I can see what
   I'd have defaulted to.

4. WHAT I'M AVOIDING
   Anything that's been sitting for a while without moving. Be direct.

Rules:
- Weigh against the goal, not against how recent or how loud something is.
- Never invent a priority I haven't committed to somewhere.
- If my goal and my task list have drifted apart, tell me plainly.
- This is a proposal. I decide.

Then write the brief with this at the top.

How to set it up

1. Write your goal down somewhere you'll actually maintain.
2. Get your commitments landing in one to-do list, however that happens.
3. Add the prioritization step to the brief you already get, rather than building a separate thing.
4. Run it for a week and see whether you agree with the proposals. Where you don't, work out whether the agent was wrong or the goal was.
5. Update the goal whenever it changes, which will be more often than you expect.

What if I only run one business?

The same problem exists at a smaller scale. One business still generates more urgent-feeling work than you can do, and the newest thing still wins by default. The goal is what stops that.

Assign

The agent reads everything, weighs it, and proposes.

I pick.

Most days I take the proposal, which is the whole point of building it. But the distinction is real, because the agent doesn't know everything. It can't see that a client conversation went badly yesterday, or that I've got a reason to move something forward early.

What it does do is make the default choice visible. When I skip past the top item, I'm now aware I'm doing it, rather than drifting into the more interesting thing without noticing.

The "what looks urgent but isn't" section is the part that earns its place. It catches the thing I'd have started with, and asks me to justify it.

Rule of thumb from this build: make the default visible, then decide.

Integrate

This is where the stack actually converges:

• One agent pulls to-dos out of the day's meeting transcripts
• One keeps the relationship records current, so anything owed to a person is visible
• One cleans the inbox, so what's left needs an answer
• This step reads all of it, weighs it against the goal, and proposes
• The morning brief arrives with the decision already in it

Ella runs at 7am. At 8am, Albus reads what she and the other agents left behind, ranks it against the goal, and sends the brief.

Four agents, one text message. Each one does a small job. Together, the day is sorted before I'm awake.

And the whole thing collapses without the goal. Everything here is measurement, and measurement needs something to measure against.

Next: the workflow that reads my analytics across every platform and tells me what's actually working.