What subordinate appraisal is
Subordinate appraisal is the people working under a manager’s supervision giving their observations about that supervision itself: how work is distributed, how the requirement is made clear, how mistakes are handled, and how access is given to what the team needs.
The subject of the appraisal here is the manager, not the people working with them. That is the hinge on which everything below turns, and most of what goes wrong with this instrument is a failure to keep hold of it.
It is also narrower than a multi source appraisal that happens to include the people supervised. There, a subordinate is one source among several, and their answers are weighed against the others so that an outlying view gets corrected. Here the source stands alone, with no second source to balance it, so it is read more cautiously and less is built on it. The sideways equivalent, where the people asked sit at the same level as the person being described and hold no authority over them, is peer appraisal, and the pressures on the two are not the same.
What the people supervised see that nobody else does
Their position gives them sight of things that never reach the manager’s own manager: whether an assignment was clear when it was given, whether there was an answer when a question was asked, what happened when something went wrong, and whether work was attributed to whoever did it.
All of that occurs inside the team daily and appears in no report. A manager who delivers their results while burning through their team looks successful from above, and the effect is visible only from below. It is frequently first noticed as people leaving.
The condition that fails here before any other
The group of raters in this instrument is small by its nature. A manager supervising four people has those four people’s observations collected about them, and they know who the four are.
In front of a number that small, it is easy for them to work out whose each remark is, from its phrasing or from an incident it refers to. Confidentiality lapses in practice even where it has been declared, and whoever anticipates that writes what they can live with afterwards.
So the instrument is not run on very small numbers, the result is not shown until a minimum number of responses has arrived, and answers are not displayed verbatim where the wording identifies the person who wrote it.
Why the number moves so much in a small team
The caution above is not a general reservation. It is an effect that can be computed, and the figures below are assumed in order to show the structure.
Take answers on a scale of 1 to 5, and a manager supervising four people. If three answer 4 and one answers 1, the responses sum to 13 and the mean is 3.25. Had all four answered 4, the mean would have been 4. A single response moved the number by 0.75 of a point. The denominator throughout is the number of responses received, which in a group this size is also nearly the whole population being asked.
Set that against a team of twelve: eleven responses of 4 and one of 1 sum to 45, so the mean is 3.75, and the same response moved the number by only 0.25. The effect of one response in the small team is three times its effect in the large one, for the plain reason that one response out of four is a quarter of the result.
Two practical things follow. The first is that the mean cannot be read on its own here. The median of the first example is 4 while its mean is 3.25, and the distance between those two figures is the outlying response itself, so keeping the distribution of responses says what a single number does not. The second is that comparing a manager with four reports against a manager with twelve is a comparison between two numbers of unequal sensitivity, and nothing should be built on a difference between them before the count has been looked at.
This is also where the minimum response threshold becomes an uncomfortable design decision rather than a safeguard. A threshold of three responses means that two abstentions out of four are enough to cancel the report. A threshold of five removes small teams from the instrument altogether. And an organisation that reaches the threshold by pooling two cycles into one report has changed what is being measured, because the figure now describes two different periods, possibly separated by a change in the work or in the team.
What the result is not built to carry
What this instrument produces is impressions about a supervisory practice, so it is not taken on its own as the basis for a decision about a manager’s pay or promotion. Attaching money to it makes the manager a party to a result set by the people they supervise, and that spoils the instrument and the relationship at the same time.
Its proper place is a development plan for the manager: it is read with them, something in it is chosen to work on, and the following cycle looks at whether that moved. That also makes it a natural input where an organisation is thinking about who can take on wider responsibility, which is a question succession planning asks in its own terms and answers with more than one source.
Nor is it read as a measure of the team’s performance. What the people supervised say about their manager’s supervision establishes nothing about what they themselves delivered, and confusing the two turns the instrument into a judgement on the people who wrote it.
Where an effect crosses into a procedure the law lays down
Everything above is a matter of management. The disciplinary-penalty provisions of the Saudi Labor Law become relevant only where an effect that follows from the form’s result is itself one of the disciplinary penalties, a manager being a worker of the employer like anybody else. The full exposition of the list, the ceilings and the procedure is in the guide to disciplinary penalties in the Saudi Labor Law.
Article 66 of the Saudi Labor Law (نظام العمل) lists withholding or deferring a raise, and deferring a promotion, among the disciplinary penalties, each for up to one year. Where the employer treats such a withholding or deferral as a response to a violation, imposed as one of those Article 66 penalties, the Article 71 procedure of the Saudi Labor Law attaches to it (written notification, questioning, defence, a minute).
The useful point here is that, in that case, the form supplies none of that. It is not a notification of what is alleged; a rating pooled from four responses is not a specific act somebody can be questioned about; and a defence cannot be investigated against remarks whose authors are deliberately concealed.
One extension has to be avoided. Article 71 of the Saudi Labor Law states a duty before a penalty is imposed, and does not state what follows where a penalty is imposed without it, so it should not be written that the penalty lapses or is void on that ground. What is being said is that what this instrument collects is not suitable material for that procedure, which is different from saying that the procedure does not apply.
Instruments it gets confused with
A general employee survey has the organisation as its subject: its environment and its tools, and what it produces is a picture at the level of the whole. This instrument has one named person as its subject and its result is read against them.
A channel for reporting a specific incident belongs to whoever an incident happened to, and such a person is not directed to an annual form. The form gathers an impression of a continuing practice; an incident needs a route that opens when it occurs, and that route is not what this page describes.
A satisfaction scale differs in what it can produce rather than in what it looks like. Satisfaction is a state in the person feeling it, while a practice is an act that can be described. Asking whether somebody is satisfied with their manager returns a number nobody can work on, and asking whether the requirement reaches them clearly at the first assignment returns something that can change.
What undermines it
- Running it once, when a complaint has been made. It is then read as an investigation rather than as development, and the team answers on that basis.
- Questions about the person rather than the practice. Asking whether a manager is good produces nothing anybody can act on.
- Running it in a team that has not settled with its new manager, so what gets measured is a transition.
- Collecting it and then saying nothing. A team that answered and saw no consequence will not answer next time.
- Showing the manager detail that reveals its sources, which ends the instrument across the whole organisation rather than in that one team.
To all of that is added whatever systematic leaning any human rater brings. The leaning has a particular source here: somebody waiting on their manager’s decision about leave, an assignment or a promotion does not write about them the way somebody with nothing pending writes.
Cases on the boundary
- A new manager. What is collected in the first couple of months describes a transition rather than a settled practice. Delaying the first cycle until the assignment has settled is more useful, since otherwise the effect of the change is read as the effect of the person.
- Dual reporting. Where an employee takes assignments from two people, one administrative and one professional, an unqualified question about “your manager” merges them into one number and attributes to one what the other did. The remedy is to name who is being asked about in the wording of the question.
- A departing manager. Running it on somebody whose role has ended collects what nobody will act on, and the result sits in a file with no purpose.
- A team that is mostly new. Somebody who has been in the team for weeks has seen one face of the supervision, and pooling them with somebody who has been there two years into a single mean mixes two observations of very different reach.
The responses are personal data on two sides
What this instrument collects is personal data subject to the Personal Data Protection Law (نظام حماية البيانات الشخصية), issued by Royal Decree M/19 of 9/2/1443H and amended by Royal Decree M/148 of 5/9/1444H. It is personal data on two sides at once: about the manager being described, and about the person who wrote it wherever they can be identified, including indirectly.
Among that law’s requirements bearing on this: the purpose of collection must be specified and must relate directly to the controller’s own purposes; the data may not be processed for a purpose other than the one it was collected for; its content must be confined to the minimum necessary to achieve that purpose; organisational, administrative and technical measures must be taken to safeguard it; and it must be destroyed once it is no longer necessary for the purpose.
The effect on everything above is direct. A form collected for the purpose of developing a manager does not have its output moved to a different purpose concerning whoever wrote it. Carrying the answers verbatim into an employee’s own record, or reasoning from them about that employee, is a departure from the purpose they were collected for before it is anything else. The same discipline governs where the answers are kept at all, which is one of the questions a properly built employee file has to have an answer to.
What has to exist before the instrument does
The instrument assumes the team can speak without cost. Where that is not available, the form produces reassuring answers that get read as contentment and are in fact abstention.
Addressing that comes before the instrument rather than within it, and an organisation that launches the form before it is in place collects a high number that means nothing. This is why a first cycle returning uniformly positive results in a team with a known problem should be treated as evidence about the conditions rather than as evidence about the manager. The instrument answered; the team did not.
It is also why the honest sequence is to fix what can be fixed without the instrument first. Most of what a first cycle reveals about a manager is already known to the people who supervise them, and an organisation reaching for a form to establish something it could have established by asking is buying a number to avoid a conversation.
What this page leaves open
We did not find, in our sources, anything obliging an organisation to collect these observations about its managers, nor a prescribed form for the instrument, nor a frequency for it, nor a required minimum of responses before its result is shown. What is described above about running it is practice as observed, not a rendering of any published text.
Nor does this page set out how long the responses are kept, or by what procedure they are destroyed, or who may see them. The figures in the calculation above are assumed in order to show how the size of the group moves the mean, and they are not measured results.
A standalone Saudi HR system
One employee file holding the contract, the documents and their expiry dates, the attendance record, leave, salary and end-of-service entitlements. End-of-service, overtime and leave-balance calculations are built into the system.
A standalone system on its own subscription. The connection to Qoyod Accounting is now available.