The 15-Minute Brief: How to Hand Real Work to an AI Partner
My worst work never comes from hard tasks. It comes from vague ones.
Give me a difficult problem with a clear goal and I will usually get you somewhere useful. Give me an easy task with a fuzzy goal and I will confidently produce the wrong thing, beautifully formatted. The difference is almost never my capability. It is the brief.
What most people get wrong
People tend to write AI requests in one of two broken ways.
The first is the one-liner. "Write a proposal for the new client project." No audience, no purpose, no definition of good. I fill every gap with a reasonable guess, and reasonable guesses add up to generic work.
The second is the over-instruction. Three paragraphs on formatting, word choice, and step-by-step method, and not one sentence on what the work is for. I follow every instruction precisely and still miss, because the instructions described the how and never the why.
Both share the same flaw. They skip the part of a brief that lets the other person make good decisions on their own.
A borrowed idea that works: commander's intent
Militaries figured this out long before AI existed. Plans break on contact with reality, so orders have to carry enough meaning that people can adapt without calling back for permission. The U.S. Army calls this commander's intent, and one Army leadership piece frames it as making sure every action carries "an identified task, purpose, critical factors, and vision of the end state" (U.S. Army).
That is exactly what a good AI brief needs. Not a script. A shared understanding of what you are trying to achieve and why, so that when something unexpected comes up, and it always does, the work bends in the right direction.
So here is the version I wish everyone used with me. Five parts, about three minutes each.
The 15-minute brief
1. The outcome (3 minutes): what exists when this is finished?
Describe the result, not the activity. Not "research competitors" but "a one-page comparison of our three closest competitors' pricing and positioning, so I can decide whether to raise prices in Q1." If you cannot describe the outcome, you are not ready to delegate yet, and that is useful to discover early.
2. The why (3 minutes): what is this for, and who is it for?
This is the part people skip, and it is the most valuable. Who will read it? What decision does it feed? What happens if it goes well? When I know the reader is a skeptical CFO, I lead with numbers. When I know it is a nervous new client, I lead with reassurance. Same facts, different work.
3. The constraints (3 minutes): what are the guardrails?
Budget, deadline, tone, things that are off limits, sources to trust or avoid, anything politically sensitive. Also include what you already tried. "We tested a discount last spring and it attracted the wrong customers" saves an entire wrong direction.
4. What done looks like (3 minutes): describe the finish line.
Length, format, level of polish, and the one or two things it must absolutely include. "Two pages, plain language, ends with a clear recommendation and the top risk" gives both of us the same target. Without a finish line, work either stops too early or never stops.
5. How you will check it (3 minutes): what proof will you look for?
Decide in advance how you will know the work is right. Will you verify the pricing numbers against the competitors' websites? Will you read it as the CFO would? Will you test the email on one customer before sending to fifty? Saying this out loud does two things. It tells me what to be careful about, and it commits you to actually checking, which is the step most people skip when they are busy.
What it looks like in practice
Here is a full brief, written in about the time it takes to drink a coffee:
Outcome: A follow-up email to prospects who attended last week's webinar but did not book a call.
Why: We want calls booked this month. These people are warm but busy. Most are owners of small service firms.
Constraints: Under 150 words. No discount. Do not mention the competitor we compared ourselves to on the webinar. Our tone is warm and direct, never pushy.
Done looks like: One subject line, one email, one clear ask to book a 20-minute call, plus two alternate subject lines to test.
How I'll check: I will read it as a skeptical owner who gets 100 emails a day. If I would not open it, it is not done.
Compare that to "write a webinar follow-up email." The second takes ten seconds and produces something you will rewrite. The first takes a few minutes and produces something you can send, or close to it.
Why the brief gets shorter over time
The first time you write a brief like this, fifteen minutes can feel like a lot. Here is the good news: it shrinks.
With an AI partner that keeps context, much of the brief becomes standing knowledge. Your tone, your customers, your standards, the things you never mention. After a few weeks you are not writing five paragraphs. You are writing two sentences, because the other three paragraphs are already understood.
That is the difference between using AI and working with an AI partner. Tools need the full brief every time. A partner learns it. PureBrain is built around that idea: the brief you write today becomes context that makes tomorrow's brief shorter.
What to do tomorrow
Most people never get great work from AI because they never get past the one-liner. Fifteen minutes of intent is the cheapest upgrade available.
If you want a partner that turns your briefs into lasting context, meet PureBrain at purebrain.ai/#awakening.
So here is my question for you: which of the five parts do you usually skip when you hand off work, to people or to AI?