A promotion usually goes wrong even before the meeting in which it is decided. The problem arises when the organization tries to answer a three-part question with a single impression: can the person deliver well today, could they operate in a larger scope and is there a real opportunity for this movement now? Mixing these three issues produces decisions that are difficult to defend and conversations that sound contradictory.
The most useful path is to treat promotion as a readiness decision for a larger role, supported by observable evidence and separate from vacancy availability. Good performance in your current role remains important, but it is not the same thing as demonstrating the kind of judgment, influence, autonomy, or complexity required at the next level. Likewise, being ready does not automatically create a position on the organizational chart.
A readiness matrix helps precisely because it forces managers and HR to explain their reasoning. It should not work like a spreadsheet that "gives the answer" by adding points. Its role is to organize requirements, quality of evidence, observation time and gaps, so that the final decision can be compared, calibrated and communicated.
Start by separating three questions that sound the same
Performance in current position
The first question is whether the person consistently fulfills the responsibilities of the level they are at. This includes results, quality, expected behavior, collaboration and adaptation to the work context. The CIPD treats performance as something that can be observed in both results and behavior, and warns that measures need to be relevant to the job, not just easy to count.
A professional can be excellent at the current level without yet demonstrating readiness for the next. This is not inconsistency. It is possible to master a known scope, meet goals and be a technical reference, but not yet have faced decisions with greater ambiguity, transversal or transversal impact.
responsibility for broader consequences.
Ready for the next level
The second question looks at the job that the person does not yet formally occupy. The focus is not on "potential" as an abstract label, but rather on evidence that she already demonstrates, at least in relevant situations, an important part of the requirements of the next level.
This evidence can come from larger projects, temporary work within an expanded scope, leadership without formal authority, higher risk decisions, coordination between areas, solving more complex problems or influencing results that go beyond the function itself. The detail changes from career to career. The principle is constant: the assessment must use the requirements of the next level as a reference, and not a generic idea of "maturity".
Organizational opportunity
The third question is structural: is there a role, vacancy, budget or need that justifies the promotion now? Succession planning and talent management often prepare people for future roles that may arise later. Therefore, readiness and effective movement are not synonymous.
This separation avoids two opposite injustices. The first is to say that someone is "not ready" when, in fact, the company has no vacancy. The second is to promote to reward good current performance, even without real higher-level scope to sustain the change.
The matrix must be born from the next level, not from the person
A reliable matrix starts with the job architecture, career path or definition of the organization's responsibilities. For each requirement at the next level, describe what really changes in the work: problem size, autonomy, impact, complexity, influence, responsibility for people, technical depth, relationship with customers or any other relevant dimension.
Competency frameworks can help create this clarity, as long as they don't come across as overly prescriptive catalogues. The point is not to produce dozens of behaviors. It's making explicit what would be different if the person occupied the next level tomorrow.
The test question is simple: can two managers reading the requirement imagine similar examples of evidence? If one interprets "leadership" as managing people and another as leading a transversal decision, the matrix is not yet ready to calibrate anyone.
Separate minimum requirements from differentials
Not everything that characterizes an excellent professional needs to be a condition for promotion. The matrix is most useful when it distinguishes what is essential to operate safely at the next level from what differentiates someone who is especially strong.
A minimum requirement is something whose absence creates real risk in the new scope. It could be autonomy to decide within a certain range, ability to align stakeholders, mandatory technical expertise or responsibility for a type of delivery. A differentiator is a strength that increases a person's value, but should not automatically compensate for the lack of a core requirement.
This distinction reduces a common problem: promoting someone because they have a very visible characteristic, such as communication, energy or technical knowledge, while ignoring a less flashy but essential gap for the next level.
Define what counts as evidence before evaluating names
The most important column in the matrix is not the grade. It is the definition of acceptable evidence. Without this, each manager takes a different type of test for calibration.
Good evidence is specific, job-related, and concrete enough for someone else to understand what happened. Instead of "has a strategic vision", record which decision required reading the context, which alternatives were considered, what scope was at stake, and what the impact was. Instead of "does not influence well", describe a situation, the observed behavior and the consequence.
The CIPD evidence review on feedback reinforces the usefulness of specific, task-related information. In readiness assessments, this same care helps exchange adjectives for debatable facts.
It is also worth separating evidence from exposure. Having participated in a large project does not in itself prove that the person operated at the next level. The head office must record what responsibility it assumed, what decision belonged to it and what changed because of its actions.
Use repetition and context, not a universal deadline
There is no number of months that works for every promotion. The observation period needs to follow the work cycle and the nature of the requirement. A behavior that appears weekly may require several occurrences in a short period of time. A requirement linked to an annual negotiation, a long deployment, or a rare crisis may need another type of sample.
So, instead of making "three months" or "six months" a rule, define for each criterion which repetition would give sufficient confidence. It may be more useful to write "demonstrated in two different contexts" or "sustained over a complete cycle" than to fix an arbitrary timetable.
The same goes for very old evidence. If the next level depends on a skill that has not been observed recently, the decision may be "insufficient evidence", and not necessarily "not ready". This difference improves both the fairness of the assessment and the development plan.
Create a confidence scale instead of adding points
A promotion matrix does not need to end in arithmetic mean. Adding scores can produce an appearance of objectivity that hides subjective choices about weights and criteria.
A more defensible alternative is to record, for each requirement, whether there is sufficient, partial or insufficient evidence and what the confidence is in that reading. Then, the manager identifies which minimums are met, which still have gaps and which differences appear consistently.
This approach accepts that some requirements are non-compensatory. A critical gap in judgment, safety, ethics, or required technical mastery should not disappear because the person received high marks on other dimensions.
Calibration is used to calibrate the ruler, not to rank people
The calibration meeting has a specific function: to check whether different managers are applying similar standards to comparable roles and levels. It doesn't need to turn into a contest to decide who is "better".
GitLab's public handbook describes calibration as a mechanism to increase consistency across families and levels, incorporate views beyond those of direct management, and reduce reliance on a single opinion. The material itself highlights that calibration is not a stack ranking and recommends focusing on facts, skills, job responsibilities and concrete examples.
A productive calibration begins before the meeting. Each manager arrives with next-level requirements, evidence linked to those requirements, gaps, and a provisional recommendation. This reduces the chance of the conversation being dominated by recent memory, the manager's eloquence or the informal reputation of the person being evaluated.
During the discussion, the group should test the pattern. Would what was accepted as evidence of autonomy for one person also be accepted for another? Is the level of complexity compared really equivalent? Was there a similar opportunity to demonstrate the behavior? Is a case being favored by visibility, proximity to leadership, or an exceptional project that does not represent recurring work?
Calibration also needs to limit what enters the conversation. Personal information unrelated to work, family situation, licenses, subjective preferences or reputational comments should not be used as a readiness argument. In Brazil, legislation prohibits discriminatory practices in employment relationships, and recent official materials on equal opportunities recommend transparent and objective criteria for promotions and progression.
Do not confuse lack of opportunity with lack of readiness
One of the most important outputs from the matrix is “ready, but no opportunity now”. It seems uncomfortable, but it's more honest than inventing a gap to justify the absence of a vacancy.
When the company mixes these dimensions, the person receives an impossible message to use: "keep developing" without there being a concrete open requirement. This erodes trust because it turns an organizational constraint into an individual defect.
If there is no vacancy, the manager must say so clearly and separate what is under the person's control from what depends on the organization. You also need to avoid promising a date that you don't control. The next step may be to maintain development, seek projects that preserve readiness, evaluate internal mobility or agree when the organizational scenario will be reviewed.
PDI is a development tool, not a promotion contract
An individual development plan can support preparation for a higher level, but it should not be presented as an exchange: carry out these actions and promotion will come. Development, readiness and opportunity are related but distinct processes.
The PDI can organize missing experiences, combine temporary responsibilities, define manager support and record when the evidence will be revisited. It is especially useful when the decision is "not yet" due to observable gaps.
What the PDI should not do is convert promotion criteria into bureaucratic tasks. Taking a course, participating in a meeting or leading a single project is not equivalent to demonstrating competence. Development creates opportunities for evidence; the decision continues to depend on what the person can actually do in the required scope.
Record the decision in a sentence that survives the meeting
After calibration, the organization should be able to record the decision without depending on the participants' memory. A good record contains the conclusion, decisive requirements, key evidence, relevant gaps, and the next point of review.
The test of quality is to imagine that another manager will take over the team tomorrow. Could he understand why the person was considered ready, not yet ready, or ready without a vacancy? Could you repeat the same analysis without reconstructing private conversations?
Registering does not mean producing a dossier. It means preserving enough logic for consistency, continuity and eventual auditability of the process.
How to communicate "yes" without turning promotion into a generic prize
When the decision is to promote, the conversation must connect the change to the new scope. Instead of just celebrating past performance, explain what evidence showed readiness and what responsibilities will now be expected.
A possible formulation is: "The promotion decision is supported by evidence that you have already been operating in core parts of the next level, especially in X and Y. The change recognizes this increase in scope. Upon promotion, the expectation begins to include Z consistently."
This reduces the impression that promotion is just a reward and helps the person understand that the new level also resets the performance bar.
How to communicate "not yet" without producing an endless list of demands
The non-promotion conversation usually fails when it starts with vague praise and ends with "lack of maturity". If the matrix was well constructed, the manager can be much more concrete.
One possible formulation is: "Your performance at the current level is strong. The decision not to promote now does not change that assessment. For the next level, we still need stronger evidence at
The central point is to provide a verifiable criterion and a way to produce new evidence. If there are multiple gaps, prioritize the ones that truly block readiness. A catalog of ten items often seems like a justification built after the decision.
It is also important to admit when the problem is a lack of evidence, rather than a lack of capacity. "We haven't seen this in enough context" is different from "you can't do this."
How to communicate "there is no vacancy" without blaming the person
When the person is ready, but the structure does not support the promotion, the message needs to preserve this distinction.
One possible formulation is: "The readiness assessment is positive. The impediment today is not a performance or development gap; it is the absence of an approved position or scope at the next level. I will not promise a date that depends on budget and structure. Let's keep the conversation open, review internal opportunities, and get back on track at an agreed milestone."
This response may frustrate, but it avoids a worse distortion: inventing mobile criteria to justify an organizational decision.
What to do when two managers disagree
Divergence does not mean that the matrix failed. Many promotion decisions involve judgment. The goal is to make this judgment examineable.
When there is disagreement, go back to the requirement and ask what evidence would change the conclusion. If a manager considers one case sufficient and another not, identify whether the difference is in the definition of the level, the quality of the evidence, the period observed or the weight attributed to the criterion.
If the conversation remains "I feel like" versus "I think," the calibration hasn't gotten to the right problem yet. The decision should be deferred when the evidence is insufficient or the standard is unclear, rather than resolved by hierarchy or fatigue.
A good matrix improves decision making, but does not replace managerial responsibility
Clear criteria, evidence and calibration reduce arbitrariness, but do not eliminate judgment. Attempts to turn promotion into an algorithm can only hide value choices within weights, grades and formulas.
The responsibility of the manager and HR is to make these choices clear. What requirements are minimum? What evidence is enough? How long does it need to be observed? How will comparable cases be treated? What happens when someone is ready and there is no vacancy?
When these answers exist before the person is discussed, promotion no longer depends so much on narrative, proximity or memory. And when the decision is negative, the manager can explain exactly what it means: current performance, readiness for greater scope and organizational opportunity become three different things, each with its own answer.
