Good Judgment Should Be Easier

I have left plenty of meetings knowing more than I did when I walked in and still being unsure what should happen next.

The report was accurate. The dashboard did what it was built to do. The conversation surfaced real issues. We still had not made a decision.

I have also helped build systems that produced exactly that result. They explained what happened, sometimes in impressive detail, then stopped just before the harder question:

What should this change?

Earlier in my career, I often treated that as a reporting problem and reached for better metrics, a cleaner dashboard, or more visibility. Those things can help, but I have become more interested in what happens after people understand the information in front of them.

Information Is Not Judgment

Information can show us what happened. It can help explain why, reveal a pattern, or challenge an assumption. What it cannot do is make a tradeoff disappear.

Someone still has to decide what matters most, what can wait, what the organization is willing to risk, and who owns what happens next.

Many of the systems we use are built on the assumption that visibility will naturally produce clarity. Give people enough information and the right decision will follow.

Sometimes it does. Other times, the added information creates another round of interpretation, another request for data, or another meeting with more people in the room.

In my experience, the issue is rarely that people are incapable of deciding. More often, they have to reconstruct too much context before they can begin. They need to work out the goal, which measure matters, what changed since the last decision, who has the authority to act, and what would cause the team to change direction.

Some of that work is necessary. The problem is when those answers already exist but are scattered across documents, systems, conversations, and people. Good judgment becomes expensive because the organization keeps asking people to rebuild the decision around the information.

Every Dashboard Makes a Promise

I think every dashboard makes an implicit promise:

After looking at this, you will understand something that helps you act.

A dashboard can keep only the first half of that promise. It can show the numbers accurately while leaving the person using it to supply the goal, the definitions, the history, the owner, and the priority.

I do not think every dashboard should prescribe an answer. That would replace judgment rather than support it.

It should reduce the avoidable work someone has to do before using their judgment. That may mean placing the goal beside the metric, distinguishing a meaningful signal from ordinary movement, or showing earlier decisions that affect the issue. Sometimes the most useful design choice is leaving out a measure that has no relationship to a real choice.

The question I ask now is:

What becomes easier after someone understands this?

If the answer is nothing, the dashboard may be reporting without helping.

More information can even make a decision harder. Each new source may arrive with a different definition, interpretation, or recommendation. A team can begin with one question and end with ten.

That is not always a problem. Complexity is real, and responsible decisions sometimes require more investigation. The distinction I care about is whether the uncertainty belongs to the decision or was created by the way the information was organized.

Sometimes the better response is stronger constraints.

A visible goal narrows the question, and a named owner makes responsibility clear. The tradeoff should be explicit. So should the time horizon and the date when the team will look again; otherwise, a temporary choice can quietly become permanent.

None of those guarantees the right decision. They make the decision easier to explain, revisit, and learn from.

What AI Changes About Judgment

AI can make information easier to find, summarize, and organize. That is real progress. Questions that once required hours of manual work can sometimes be explored in minutes.

Faster retrieval does not settle a tradeoff.

An AI system can summarize five reports or generate ten recommendations. The organization still has to decide what deserves its time, money, and attention. Ten recommendations do not help much if nobody can explain which one should go first.

As analysis becomes easier to produce, choosing what matters requires more of our attention.

I am more interested in AI that helps people preserve context, make assumptions visible, see tradeoffs, and focus their judgment where it adds value.

AI can shorten the path to a decision, but the responsibility for making it still belongs to a person.

Design for the Next Decision

A great deal of analytics work is designed around producing an insight. I am more interested in the moment immediately after it.

A team learns that a landing page is losing customers, a product feature is not being adopted, or a process is making people repeat the same work every week.

Now what?

The insight can enter a list no one owns, become a recommendation separated from delivery, or disappear into a presentation after the meeting. Or it can remain connected to the goal, the choice, the owner, the action, and eventually the outcome.

Signal. Context. Choice. Action. Learning.

The sequence matters because an insight only improves something when it reaches a choice, an owner, and an action.

This changes what I ask of a system. Does it make the important thing visible and clarify who owns the next move? Can the team distinguish what is urgent from what is merely loud? Does the system create room for judgment, or consume it?

Those questions apply beyond software. I have come to see a meeting, a planning process, a product roadmap, or an executive brief as a decision system. Each can shorten the distance between evidence and action, or add another layer people have to navigate.

I am not trying to make decisions automatic. A useful system exposes the real choice and leaves it with the right person.

Judgment has to remain human because reality changes. Customers surprise us, constraints move, and a strategy that once made sense can stop making sense. The same signal can require a different response depending on timing, risk, and context.

People still have to understand another person’s needs, recognize when the stated problem is not the real one, balance outcomes that do not fit neatly into a metric, and accept responsibility for a tradeoff.

Finding the latest spreadsheet does not belong on that list.

A team can be uncertain and still have clarity. It can be honest about what it knows, what it does not, the choice it is making now, who owns it, and when it will look again.

That is often enough to move. It also gives the team something useful to examine when the outcome arrives.

I want to build products, workflows, and internal systems that clear away avoidable work so people can give their attention to the choices that actually require it.

Good judgment will never be effortless. A person still has to choose. I just do not think they should have to reconstruct all of the surrounding context first.