How to Turn a Vague Idea Into a Useful Brief

September 13, 2026 · By Michel Valenzuela, Founder of Daryl

“I want to build something that makes this easier.”

That is often where an idea begins.

You recognise a frustrating process, imagine a better experience, or notice a possibility others seem to overlook. You can feel the direction before you can explain it.

Then someone asks what you want to make.

You describe features, give examples, change the explanation halfway through, and realise that several different ideas are competing for the same sentence.

A useful next step is to develop a brief: a clear account of the problem, the people involved, and the outcome you want to create.

AI can help you work toward that clarity.

Begin with the moment that prompted the idea

Before describing the solution, explain what happened.

Perhaps you had to repeat information across several tools. Perhaps a client struggled to understand a process. Perhaps you found yourself doing the same manual task every week.

Describe the situation in concrete terms.

“After each client meeting, I spend time reconstructing what we agreed, identifying unanswered questions, and deciding what belongs in the follow-up.”

That gives the discussion something observable.

“I want an intelligent productivity platform” leaves much more room for interpretation. The assistant may respond with a broad feature list before you have established which difficulty deserves attention.

The original moment helps keep the idea grounded.

Identify who experiences the problem

You may understand the difficulty because you experience it yourself. Other people may encounter it differently.

An independent consultant, a project manager, and a client could all participate in the same meeting while needing different things afterward.

Choose whose experience you are trying to improve first.

Ask:

“Based on this situation, help me distinguish the different people affected. What does each person need, and whose problem am I proposing to solve?”

Treat the response as a set of possibilities to examine. The assistant cannot establish what people actually need without evidence.

Its value here is helping you recognise distinctions worth investigating.

Define the change you want to produce

What should become easier, clearer, faster, or more dependable?

For the meeting example, the desired outcome might be:

“Leave each meeting with an accurate record of commitments and a clear next action.”

That is more useful than immediately specifying an interface, an integration, or a collection of automated features.

Once the outcome is clear, you can compare different ways to achieve it.

A template might help. A better meeting process might help. Software might help. Some combination may be appropriate.

The brief should preserve room to discover the approach.

Separate observations from assumptions

An idea usually contains both.

You know that reconstructing your own meeting notes takes time. You may assume that other consultants experience the same difficulty, that they would trust an automated summary, or that they would pay for a tool addressing it.

Those assumptions could be reasonable. They still need checking.

Ask the assistant:

“Separate what I have observed from what I am assuming. Which assumptions would need to be tested before committing substantial work?”

Then correct its interpretation.

A polished brief can make an untested idea sound established. Keeping the assumptions visible prevents the writing from outrunning the evidence.

Give the first version boundaries

An idea becomes difficult to act on when it tries to address every related problem.

The meeting example could expand into scheduling, recording, task management, client communications, project reporting, and financial forecasting.

Each addition may have a plausible connection. Together, they can obscure the original purpose.

Define what the first version needs to accomplish.

You might say:

“For the initial version, focus on reviewing commitments and unresolved questions after a meeting. Scheduling and sending messages are outside the scope.”

Those boundaries let you assess proposed features against a specific objective.

You can revisit them later with a reason.

Ask for a brief you can challenge

Once the main elements are available, ask the assistant to organise them.

“Turn this discussion into a concise brief covering the user, the situation, the problem, the intended outcome, the initial scope, and the assumptions we need to test. Preserve unresolved questions.”

Then read it as if someone else had proposed the idea.

Does it describe the problem you actually care about? Has it introduced claims you cannot support? Does the scope fit your resources? Could another person understand what you want to investigate?

Edit until the brief reflects your intention accurately.

Use the brief to guide a real conversation

A brief gives you a way to ask better questions.

You can show the situation to someone who experiences it, discuss their current approach, or try a small version of the proposed improvement.

Their response may challenge the idea.

Perhaps the difficult part happens before the meeting. Perhaps the existing process works well enough. Perhaps accuracy matters much more than speed.

That information should change the brief when it changes your understanding.

The document is useful because it helps the idea develop through evidence and discussion.

Where Daryl fits

Daryl is a personal AI for Mac, currently in development, designed around local conversation and support for decisions.

I want it to provide an approachable place to bring ideas that are still taking shape: explain the situation, explore interpretations, and work toward something clear enough to examine.

The workflow described here is an illustration of thinking with AI, rather than a demonstration of a released Daryl feature.

As a founder and creative practitioner, I value the point where an intuition becomes an idea another person can understand.

A useful brief helps you reach that point without pretending every question has been answered.

You can then decide what deserves to be built, tested, or reconsidered.