A Seven Day Plan to Get Morning Brief Automation Working

Plenty of people doubt that morning brief automation works outside a demo. That is a sensible doubt. The honest answer is that it depends on how carefully you set it up and check it, and a week is enough time to find out whether it works for you.

This post is not a diary of my results, and I am not going to give you numbers I did not measure. Instead it is a seven day plan, with an illustrative example running through it. The example is a made up business, clearly labelled, so you can see what each day looks like and run your own version.

The lesson behind this is Lesson 6.5: A short brief waiting every morning. It is listed as coming soon, and this plan works as a practice run until then.

In one week you can build a small brief, check it against reality, and decide with evidence whether to keep it.

The example we will follow

Say you run a small online shop selling handmade candles, with one part time helper. This is invented for illustration. Your mornings involve email, direct messages, your order list and a calendar with a few supplier calls.

The goal for the week is modest: a short note each morning that lists new orders, messages that need an answer, and today's calls. Not more.

Day 1: write the note by hand

Do not touch any tool today. Open yesterday's email, messages and orders, and write the note you wish you had received this morning. Keep it under 150 words.

For the candle shop it might read: "Quiet day. Three new orders. One customer asks about a late delivery and needs a reply. Supplier call at 11." That is your target. Everything after this is just getting a tool to produce something like it.

What can go wrong today: writing too much. If your hand written note is long, shorten it. Government plain language guidance recommends short paragraphs with one topic each, and the same applies here.

Day 2: pick sources and check access

List where each part of the note comes from. Orders come from the shop, messages from inbox and chat, calls from the calendar.

Today, also try the easy built in option. Google Calendar can send a daily agenda by email each morning, so turn that on and see it arrive tomorrow. For the other sources, decide what an assistant is allowed to read, and look at the permissions before you accept anything.

What can go wrong today: giving access to more than you need. Connect only the sources that your note uses.

Day 3: first run, with a safety rule

Give an AI assistant your hand written note as the example. Tell it what to include, and add three rules: use only the information provided, say when something is unclear, and never send or reply.

Run it once, on yesterday's data, and compare the result with your Day 1 note. Differences are the useful part. Note each one.

What can go wrong today: invented detail. If the assistant adds something that was not in the data, tighten the first rule and name what happened.

Day 4: tidy the input

If email is part of your note, tidy it. Gmail filters can move mail to a label or archive it, and they only apply to messages that arrive after you create them. Filter out receipts and newsletters so the assistant has less noise.

Run the brief again and see whether it improves. A cleaner inbox often does more for the note than a better instruction.

What can go wrong today: a filter hiding something important. Check the filtered label once at the end of the day.

Day 5: schedule it

Pick a delivery time before you start work, and a place you open without thinking. Set it up and let it arrive tomorrow morning on its own.

What can go wrong today: scheduling it for the time you sit down. If it arrives after you are already in your inbox, you will not use it.

Day 6: the first real morning

Read the note first, before opening anything else. Then check it against the sources. Ask three questions:

• Did it miss anything that mattered?
• Was anything in it wrong?
• Did I use every line?

Write down the answers. This is the most informative morning of the week.

Day 7: review and decide

1. Count what you usedLook at the lines you acted on and the lines you skipped.
2. Count the missesNote any item you had to find yourself.
3. Count the errorsNote anything the brief stated that was wrong.
4. Make one changeFix the biggest problem by changing the instruction.
5. DecideKeep going, change course, or stop. All three are fine.

How to judge success without made up numbers

You do not need a statistic. You need three simple answers from your own week:

• Did the note arrive when expected?
• Did I trust it enough to act on it?
• Did it change how I started the day?

If you answer yes to the first two and no to the third, the brief works but has not yet earned its place. Cut it down and see whether it becomes more useful.

What an honest result might look like

A possible outcome, again illustrative, is a brief that gets two of the three items right and needs more instruction for the third. That is a normal first week. The point is that you found the gap in seven days, not seven weeks.

Another is that you decide the manual way suits you better, perhaps because your mornings change too much. That is a legitimate result too, and it is far cheaper to learn it in a week.

Common questions

Questions people ask

Can I really do this in a week?

A basic version, yes. A polished one takes longer, because you improve it by using it.

How much time does each day need?

Plan for under an hour on the setup days and a few minutes on the checking days. These are rough planning figures, not measured results.

What if I get stuck on a day?

Skip ahead to Day 6 with what you have. A rough brief you read is better than a perfect one you never run.

Do I need to pay for anything?

Not necessarily. Check what your current tools already include before buying anything.

After the week

If it worked, add one item at a time and keep the Friday review. If you want the guided version, follow Lesson 6.5, A short brief waiting every morning. For ideas on what to add next, try 12 examples you can adapt.