The leadership team is reviewing a supervisor development program.
The presentation begins well. One hundred eighty-four supervisors attended. Ninety-six percent completed the program. The average participant rating was 4.8 out of 5. All six modules were delivered on schedule.
Those are good numbers. They show that people came, the program was delivered, and most participants valued the experience.
Then someone asks:
“Are the supervisors actually handling performance conversations better?”
The room gets quieter.
Nobody has necessarily done anything wrong. The training team has data. The program was completed. Participants were satisfied. Every promised activity happened.
But the question has moved beyond delivery.
It is asking whether the work made contact with the reason the work existed in the first place.
Progress reports are easier to produce
Most organizations know how to report activity.
We can count participants, meetings, emails, calls, pages, sessions, transactions, milestones, downloads, hours, and completed deliverables. These numbers are usually available because the activity happened inside systems we control.
Results are more difficult.
You can promise to hold the workshop. You cannot command every supervisor to use what they learned afterward. You can launch the customer portal, but you cannot guarantee that customers will find it easier to use. You can introduce a new process without controlling every condition that determines whether people follow it.
So when someone asks, “How did it go?”, we naturally return to what we can defend.
We completed everything.
We reached the target number.
We delivered on time.
That matters.
It is simply not the whole answer.
A progress report tells us what we did. Proof helps us decide whether it worked.
Delivery deserves to be measured
It would be a mistake to swing too far in the other direction and dismiss activity measures as meaningless.
If you promised to train 200 supervisors and only 63 attended, that matters. If the new system was supposed to launch in June and it still does not exist in August, that matters. If the customer team agreed to contact every affected client and half were never reached, you have a delivery problem before you have an impact problem.
Execution deserves accountability.
The distinction is that delivery measures answer one kind of question:
Did we do what we said we would do?
Proof answers another:
Did what we did begin producing the change we were trying to create?
You need both.
The danger comes when evidence of delivery quietly becomes evidence of success.
One hundred percent completion proves that the program was completed.
It does not yet prove that anybody works differently because of it.
The work may be finished before the result is visible
This is why Ask What Became Different Because You Worked asks you to return to reality after the project ends.
Some results appear quickly.
If you rewrite a confusing customer instruction, you may know within days whether people can complete the next step more easily.
Other results need time.
A supervisor can practise a feedback conversation this afternoon. Whether that practice becomes part of how they actually lead may take several weeks to observe. Whether the change contributes to lower rework, earlier problem solving, or stronger performance may take longer still.
Demanding final results too soon creates bad conclusions.
Never asking about them creates another kind of bad conclusion.
So proof needs a time horizon.
You need to know what you should reasonably expect to see now, what should become visible later, and what would be premature to claim.
Look for proof at the distance the work has traveled
Suppose the goal of a leadership program is to help supervisors handle difficult performance conversations rather than immediately passing them upward to managers.
On the day of training, one useful question is:
Can the supervisor perform the conversation?
You can watch the practice. Listen to the language. See whether the supervisor can make the expectation clear, respond to resistance, and close with a useful next move.
That is immediate evidence.
Two weeks later, the question changes:
Did the supervisor actually use the conversation at work?
Now you are looking for behavior. Did the play leave the classroom? Did supervisors attempt it during real situations? What happened when they did?
Later still, the question can move farther:
What is happening to the recurring problem the conversation was meant to address?
Are issues being handled earlier? Are managers receiving fewer avoidable escalations? Is rework declining in situations where clearer conversations matter?
Those are different distances from the original intervention.
Do not demand the third kind of proof while people are still producing the first.
But do not celebrate the first forever if the reason for doing the work was the third.
Decide what you will look for before you know whether you succeeded
This is where reporting can become slippery without anyone deliberately trying to deceive.
Imagine that a new program is launched to reduce repeated customer calls.
Three months later, repeated calls have not declined.
But satisfaction scores from the launch event were high.
So the final report emphasizes satisfaction.
That number may be real. It may even tell you something useful.
It does not answer the question the work was originally created to answer.
Humans are very good at protecting work we have already invested in. Once time, money, reputation, and effort are involved, we naturally become interested in the evidence that makes the project look worthwhile.
That is why one of the most important proof questions should be asked before the result arrives:
If this works, what should we reasonably be able to see—and when?
You may later discover that the original measure was wrong. Reality can teach you that too.
But deciding in advance makes it harder to transform whatever number looks best into the definition of success.
Proof is not a verdict on the people who did the work
There is another reason teams retreat into activity reports.
Proof can feel dangerous.
If customer behavior did not change, somebody may blame the project team. If supervisors are not using the practice, the facilitator may fear that the program will be declared a failure. If the new system did not reduce delays, the people who designed it may feel that the evidence is judging their competence.
So the report becomes defensive.
We completed all activities.
We achieved 98 percent participation.
Participants rated us highly.
Sometimes those facts are presented not because they answer the real question, but because they are safer than saying:
We do not yet see the change we expected.
That sentence can feel uncomfortable.
It can also be extremely useful.
Proof should not be treated as a verdict on your worth. It is feedback on the bet you made.
You believed that doing this work would help create a particular result. Now reality is giving you information about that belief.
Perhaps the work was strong but the environment prevented people from using it.
Perhaps one part of the intervention worked and another did not.
Perhaps the result needs more time.
Perhaps you solved the wrong problem.
Perhaps the work genuinely needs to change.
The evidence becomes valuable when people are safe enough to learn from it.
Do not demand numbers where better evidence exists
Proof does not always arrive as a dashboard.
A manager observing supervisors handle conversations differently can be evidence. Repeated customer interviews can reveal that an instruction is now clearer. A pattern across documented cases can show that people are making decisions they previously escalated.
Numbers are especially useful when frequency, scale, trend, or comparison matters. But forcing every human result into a number can create false precision.
What matters is whether the evidence is credible enough for the decision you are trying to make.
If you are deciding whether to test an idea for another month, a handful of well-observed cases may be enough.
If you are deciding whether to spend millions rolling the intervention across an organization, you probably need stronger evidence.
Proof should fit the consequence of the decision.
Close with evidence
When meaningful work reaches a review point, move beyond the usual list of activities.
Start with what you promised to deliver. This protects execution accountability. Did the work actually happen as intended?
Then return to the change you were trying to create. Why did anyone agree to spend time, money, or attention on this work in the first place?
Bring forward what you can actually see now. Use the strongest evidence available at the appropriate distance from the work.
Then say what you cannot yet claim. Perhaps the time horizon is too short. Perhaps attribution is uncertain. Perhaps one part of the evidence is still missing.
Finally, answer the question that gives proof its real purpose:
What does this evidence tell us to do next?
Continue?
Adjust?
Run another test?
Scale?
Stop?
Investigate a different part of the problem?
That final question changes the review from a report into a decision.
A stronger closing conversation
Imagine the leadership program review again.
Instead of ending with attendance and satisfaction alone, the team says:
“We delivered all six sessions to 184 supervisors, with 96 percent completion. The program was intended to help supervisors handle performance conversations before issues required manager intervention.
“During practice, most participants demonstrated the core conversation successfully. Over the next three weeks, 61 supervisors reported using the approach in a real situation, and several managers reported seeing earlier handling of performance issues. We do not yet have enough evidence to say whether escalations or rework have declined across the organization.
“Our recommendation is to continue the practice for another six weeks, track actual use more consistently, and compare escalation patterns in the teams where supervisors are applying it.”
That report is less triumphant than “The program was a great success.”
It is also more useful.
A decision-maker now knows what happened, what appears promising, what remains uncertain, and what the team recommends doing next.
The evidence is doing work.
Do not ask proof to flatter the project
A project review should not be a courtroom where the team must prove its innocence.
Nor should it be an awards ceremony where every completed project needs to become a success story.
Sometimes the evidence says, “Continue.”
Sometimes it says, “Not yet.”
Sometimes it says, “This part worked, but the larger condition has not moved.”
Sometimes it says, “Stop putting more effort here.”
That last answer can be painful, especially after good people have worked hard.
But work does not become more meaningful because we protect it from reality.
It becomes more meaningful when reality helps us make the next move better.
The purpose of proof is the next decision
You have already finished the task. You have already gone back to see what became different. Perhaps you have also made the contribution visible so other people understand what changed and what part you played.
At the review table, the job changes again.
Do not bring only evidence that people worked.
Bring enough evidence to help people think.
What did we deliver?
What change were we trying to create?
What can we see?
What remains uncertain?
What should happen next?
Then allow the evidence to answer without forcing it to praise you.
Do not ask proof to flatter the work. Ask it to guide the next decision.
That completes the first set of Make Your Work Matter articles. The next article looks at a different tension: what standard do you choose when nobody is watching and recognition may never arrive? Continue with Excellence Is a Choice You Make Before Anyone Notices.