A Decision Needs a Next Step

September 10, 2026 · By Michel Valenzuela, Founder of Daryl

You have spent an hour thinking through a problem.

The options are clearer. You understand the trade-offs. You know which direction makes sense.

Then you close the conversation and return to your day.

By tomorrow, the decision is still sitting there. You agreed with the conclusion, but nothing has changed.

This is where a useful discussion needs one more piece: a practical next step.

I want that connection between understanding and action to be part of how Daryl helps people think.

A direction still needs to become something you can do

“Reduce the project scope” is a direction.

“Ask the client whether the first phase can cover the two essential deliverables” is an action.

The difference matters because the second statement tells you how to begin.

Many decisions stop at the level of intention:

Improve the process. Protect more time. Explore a different approach. Clarify expectations.

Those intentions may be sound. They still leave you to work out what happens when you open your calendar or return to the document.

A next step should make the first move visible.

Start with what the decision actually requires

Suppose you have decided that a project needs a later deadline.

You may need to check a dependency before proposing a date. You may need a conversation with a colleague. You may already have enough information to write to the client.

Those are different starting points.

Ask:

“Given this decision, what needs to happen first? Identify anything I must confirm before making a commitment.”

An AI can help you examine that sequence. Check its suggestion against the circumstances you know.

The first action should address the real dependency. Otherwise, you can spend time producing work that has to be redone after an unresolved detail changes.

Make the action small enough to begin

A next step can become another project if it is too broad.

“Develop a new business strategy” offers little help when you have thirty minutes available.

“List the three services that produced the strongest results this year” gives you a manageable starting point. That information can then support the next discussion.

Small actions are useful when they produce something relevant: a confirmed fact, a draft, a comparison, or a conversation that resolves uncertainty.

The point is to advance the decision with the time and resources you have.

Describe what completion looks like

A clear action needs a stopping point.

If your next step is “research the options,” how will you know when you have done enough?

You might define it as:

“Find three candidates that meet the location and capacity requirements, then record the source for each.”

Now the activity has an output you can review.

Or, for a scope discussion:

“Prepare two delivery options, each with a defined scope and proposed date, so the client can choose.”

That gives the work a purpose and makes completion easier to recognise.

Know who is responsible

Some next steps belong to you. Others depend on someone else.

“Confirm the deadline” is incomplete if nobody knows who will ask or who can approve the answer.

For collaborative work, specify responsibility:

“Ask the project lead to confirm the final review date.”

If you are the person making the request, decide when you will send it and what you need back.

This is especially useful after meetings. A shared conclusion can disappear into the space between several people assuming someone else will act.

A practical next step makes that responsibility explicit.

Let the next action test your assumptions

You do not always need to commit to the entire plan.

If the decision depends on whether a client will accept a smaller scope, the next step can be to ask. If an idea depends on a workflow being useful, the next step can be to try a limited version.

A focused action can produce the information needed for a better decision.

You might ask an assistant:

“What is a practical way to check the assumption this plan depends on?”

Then assess the effort and consequences involved.

Some tests are easy to reverse. Others create expectations or commitments. Choose a next step appropriate to the uncertainty you are trying to resolve.

Keep preparation and execution distinct

An AI can help you draft a message, organise a plan, or identify the next action.

Carrying out that action is a separate step.

You may want to edit the wording, confirm a figure, or reconsider the timing before sending anything. Preparing a useful draft should leave room for that review.

That distinction is central to the experience I want to build with Daryl. Helping someone reach an actionable conclusion should preserve their control over what happens afterward.

Revisit the decision when the facts change

The next step may reveal that your chosen direction needs revision.

A deadline cannot move. A supplier has capacity after all. A client explains a requirement you had misunderstood.

Keep the reason for the original decision available so you can judge whether the new information changes it.

You can ask:

“This is what I learned. Does it change the comparison we made, or only how we carry out the decision?”

That keeps the action connected to the thinking that justified it.

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 help people move from a situation they are trying to understand toward a next step they can assess and choose.

The examples here describe that product direction. They are not claims of released task execution or autonomous action.

A useful conversation can leave you with a clearer question, a revised option, or a reasoned choice.

It should also help you understand what to do when the conversation ends.

That is where the decision begins to change the work.