Who is next? Justify service priority before conflicts arise
Choose between arrival order, urgency and purchased preference, and test how your service resolves a conflict between them.
Article language: English

In this article
At 8:10, a hypothetical company receives a question about changing a label layout for next month. At 8:40, another customer reports that they cannot prepare labels for today's shipments. At 8:45, the account manager for a major customer calls: "Mine is urgent too." Three requests, one person available to handle them. Writing URGENT in capital letters will not decide who should be next.
The company supports users of a labelling system. We want to choose a rule for access to the same support resource: order of arrival, priority based on the consequences of waiting, or preferential treatment purchased as part of an offer. Each has a different justification. Mix them at random and the real rule becomes how effectively a customer can reach an employee they know.
The same colour can represent different commitments
The owner should first decide what priority changes. It could mean an earlier assessment of the request, an earlier start to the work, or access to a separate team. Those are different promises. A customer entitled to a faster assessment does not automatically receive an earlier solution, especially if the case needs more information.
Order of arrival is easy to explain when requests are comparable: the earlier request gets access to support first. But the company needs to define when a case enters the queue. Is it the first message, acceptance of the request, or delivery of the material needed to start work? A customer should not lose their place because of a rule they only learn about days later.
Urgency, by contrast, comes from what happens while the customer waits. In our example, shipments being held up today may justify an earlier assessment than a planned label redesign. First, establish with the customer the scope of the problem, a possible workaround and when the consequences will begin to matter.
Zendesk describes its own support criteria in terms of business impact, the availability of a workaround and the extent to which work is affected. It also describes reassessing impact as a request is handled. This is one provider's example, not a ready-made standard every company must copy along with its level names.
Purchased preference rests on a different basis: the company has made a particular commitment to the customer. Do not quietly call it urgency of impact. A question about a future change may have priority within a specified service even though nobody's work has stopped. The team should know what to do with that preference when a genuine stoppage appears alongside it.
Resolve the conflict using actual cases
Return to the morning. The first message contains a complete description of the new label layout. The second reports that label preparation has stopped, but does not explain whether the customer can continue another way. The third comes from a customer who has bought faster support and concerns preparation for a future demonstration. If the company sorts solely by arrival time, it starts with the first. If it sorts by package, it may start with the third. If it sorts by impact, it should clarify the second first.
A quick clarification does not mean abandoning every other case for the rest of the day. The person assessing it can find out whether current shipments are actually affected and whether there is a useful workaround. Only then do they choose what to do next. In our example, a case whose impact needs checking can therefore receive priority for a short assessment before it receives priority for the full work.
Suppose the customer can use a second workstation and continue, albeit more slowly. That changes what we know about urgency, but it does not remove the problem. The company should be able to change the request's place once the facts are established and say when it will return to it. A label assigned at entry need not be a lifelong property of the case.
Now reverse the situation: while handling the first customer, the specialist learns that their "future change" is needed for a new order shipping that same day. The original description did not convey the consequence of waiting. The rule should allow a reassessment without forcing the customer to submit a new request. Otherwise, the customer may learn to begin every message with the highest possible alarm.
For that decision, a clear description of the grounds and someone authorised to apply them are enough. There is no need to assign points to every word. Two cases with similar impact can still fall back on arrival order as a tie-breaker. When significant uncertainty remains, it is better to name it than hide it behind an apparently precise number.
Suppose the specialist is already speaking with the first customer when the report of halted shipments arrives. Priority for the new case may mean the next available slot. Interrupting work already under way needs a separate decision: can someone else assess the impact, can the conversation reasonably be finished, and what will we tell the person whose work we are interrupting? Moving a card to the top of a board does not answer those questions.
The owner may designate someone to assess new urgent situations while retaining a different order for full resolution. But they must check whether that person has the necessary knowledge and can communicate the decision. Making them only a recipient of messages will not resolve the conflict over access to the specialist. In our company, the short assessment should produce a concrete next action, not just another note saying "urgent".
This is a different task from setting priorities across whole projects (in Polish) (in Polish). Here you are changing the order in which customers access support already offered. The decision should be explainable to the person receiving the next request without renegotiating the company's strategy.
Purchased priority needs a defined scope
An offer of "priority support" is incomplete if the company does not know which queue it provides preference over. Does the customer move ahead of ordinary questions? Do they have a separate period of availability? Does priority cover every topic or selected kinds of problem? The answers should match what the team can actually deliver when several such requests arrive at once.
In the abstract of their 2024 paper, Frazelle and Katok describe a paid-priority model and laboratory experiments. Some participants' observed choices departed from the strategies predicted by the model, including by buying priority more often than those strategies suggested. The finding does not allow the purchase itself to be treated as a perfect measure of a customer's cost of waiting.
In our fictional company, a commercial preference can be explicitly separate from the impact of a problem. The company then needs to decide how those two pieces of information jointly determine order. Different service options could instead have separate resources. No arrangement, however, creates another person just because the word "priority" appears in an offer.

Pay particular attention to promises made to several customers at once. If each can "always be first", the team receives conflicting instructions. A rule is needed within the preferred group too. It should cover the point of entry and situations where a new case calls for a faster assessment of its impact.
Before changing the offer, review actual incoming work and the availability of the right skills. You cannot establish a workable standard from the number of messages alone. Some cases end with information; others wait for a person who knows a particular solution. The preference has to relate to the resource the customer actually needs to access.
The queue also exists outside the ticket view
If a form preserves order but a phone call always interrupts the work, the company has an additional rule. If a message to an account manager bypasses ordinary intake, it has another. Look at these paths together. A person handling requests may keep the board in good order while someone outside it allocates most of their time.
In his 1985 essay on waiting, David Maister draws attention to a phone call receiving priority over a customer physically present. He treats it as a question of perceived fairness in the queue.
In our company, a call can serve to understand the situation quickly, but choosing that channel does not necessarily have to change the position for full resolution. Decide how the information enters the shared case. A customer who used the form should not discover that the only effective way to speed things up is to repeat everything through three other channels.
Also distinguish a follow-up contact from a new arrival. A further message may contain information that materially changes urgency. It may instead ask about the status of the same case. The first calls for assessment of a new fact; the second calls for an update about the wait. A contact count alone is not a sufficient reason for priority.
When handing over a shift and ongoing cases (in Polish) (in Polish), preserve the reason for the preference. "Handle this first" without the grounds can remain in force long after the situation changes. The next person should know what has been confirmed, what we are waiting for and when to reassess the order. A change of employee should not accidentally reset the customer's waiting time.
Aeneia Audit: When tools or AI help route requests, let us examine what data they use to establish priority and where a person takes over again. We will start with the actual flow of information and examples of changes in priority. Review the tools supporting your service queue.
Explain the rule without promising everyone will agree
A customer who waits longer may need an explanation of why the order changed. That does not require disclosing other customers' data. You can describe the kind of situation given priority and say what will happen to their case. Avoid reassuring them with "soon" alone if the company has no basis for that estimate.
In the abstract of their study published online in November 2024, Althenayyan and colleagues describe three experiments comparing paid priority with one person taking another person's place in a queue. Even when waiting time was held constant, an express line was judged less fair than a one-to-one substitution. The way an order changes can therefore affect how it is perceived independently of the minutes themselves.
Our company needs to explain its own choice and see who bears the burden. If new cases displace the same ordinary questions every day, making the rule visible will not by itself ensure those questions get handled. The owner should decide how the team will return to longer-waiting cases and who will notice when the chosen standard is exceeded.
Look at the oldest request. Is it still waiting for work, or for material from the customer? Has its impact changed? Can the employee see it in their view at all? Protection against ordinary cases being overlooked has to lead to action, not just another red mark in a report.
In rules for booking and releasing a place (in Polish) (in Polish), a similar choice appears when capacity becomes available again: who may use it next? Here the same consequence concerns a specialist's time. Allocating it to one case changes the chance of starting another, even if both fit in a single spreadsheet.
Test the rule before the next urgent call
Take several completed requests and arrange them as they appeared on arrival. Remove knowledge gained only later. Ask two people who support customers to identify the next case under the proposed rule. Differences will show where a criterion is unclear or needs information the company does not have at intake.
Then introduce a change: a workaround found, an earlier deadline, or a new request from a customer with preference. Check whether the team can update the order and preserve the reason. If the outcome changes only when the customer's name is revealed, check whether the official rule really describes how decisions are made.
Do not assess the exercise by the speed of the first message alone. Response time and time to resolve a request (in Polish) (in Polish) describe different stages. A case may receive a number quickly and then have nobody assigned to work on it. Priority should refer to the event the company actually promises to accelerate.
Also look at too much work started at once (in Polish) (in Polish). A rule for choosing the next request will not solve the problem if every new preference starts another case while the previous ones stay open. Changing a place in the queue and deciding to interrupt work already in progress should be recognisable as separate actions.
At Aeneia Audit, we can examine the role of tools and AI in that flow: impact data, routing and checks on incorrect classifications. The company sets the rule for access to its service. A tool should apply that rule to verifiable information rather than replace an organisational choice with a default sort on one field.
At 8:45 the account manager may still call. The employee should already know which new fact to ask about, who resolves the conflict and what the decision will change for other customers. Then "urgent" becomes the beginning of an explanation, not a password for a side entrance to every queue.







