The Service Queue Is Growing. Hiring May Be the Wrong First Fix.

Requests are waiting longer, and the manager asks for more service-desk staff. The proposal is understandable: there is more unfinished work than the team seems able to handle, so additional people look like the direct response.

Then the requests are examined more closely. Much of the delay happens after the first reply, while cases wait for missing information or a decision from the team that must complete the service. More people answering new requests might move work into that same queue faster without helping it leave.

In this hypothetical service, I would not reject the staffing proposal. I would change the question it needs to answer: where would an additional person remove waiting that customers actually experience? That allows the investigation to support hiring where it helps, instead of treating hiring and process improvement as opposing beliefs.

Follow one request beyond the first reply

A service desk can acknowledge a request, classify it, and send it onward while the employee still lacks the application access they need. The first team has completed its step. The service has not yet supplied the result.

For an illustrative access request, the next team might need confirmation of the appropriate application role. If that information is missing, a resolver may send the case back. Another reply then arrives, but no usable decision travels with it. A request can accumulate handling activity while spending most of its life waiting between those actions.

Timestamps help locate the delay. Reading the actual exchanges helps explain it. Perhaps nobody obtained the approval, perhaps the necessary information was not passed on, or perhaps a fully prepared request reached a team with insufficient capacity. These explanations can produce a similar queue count while supporting different investments.

Following work through its stages is the purpose of value-stream mapping. NIST's published manufacturing examples illustrate how mapping can expose a constraint and lead to a different intervention than simply adding effort at the most visible step. They are background for the approach, not evidence that this fictional service has the same cause or result. [1][2]

Better information is worth trying when information is what stops the work

Suppose the handoff appears to be the problem. A focused change to one request type can test that explanation without redesigning the entire support operation. The receiving team identifies the information it needs, and the service supplies it through the part of the process that can reasonably obtain it.

That last detail matters. The customer may know which application they need and what work they are trying to do, but not the internal permission model. Adding a mandatory “required application role” field could turn the missing information into a guess. The relevant context might instead come from an approved role mapping or a person who understands the authorization decision.

The improvement should be judged by whether the next team can act and the request finishes sooner, not just by whether the form has fewer empty fields. Comparing similar requests and noting other changes gives the result a more credible explanation. There is little benefit in changing intake, staffing, and approvals together and then pretending the outcome identifies which was responsible.

The result can strengthen the case for hiring

Now consider a possible trial outcome, not an observed result: requests arrive with the necessary information, yet the fulfillment queue still grows. That would weaken the explanation that missing context accounts for the remaining delay. The next question is what prevents complete, approved requests from being handled.

If the receiving team lacks enough qualified people for that work, additional capacity there may have a stronger case than more front-line capacity. If requests still wait for an absent approver, hiring another resolver may leave the decision untouched. The investigation has improved the staffing conversation by locating the work an additional person would actually perform.

That distinction also prevents an attractive local result from becoming a false success. A smaller desk queue can coexist with a longer fulfillment queue. Fewer submissions can reflect customers abandoning a difficult form. Completion time, repeat contact, and customer effort help show whether the waiting disappeared or merely moved.

The original request for staff may eventually be approved, redirected, or replaced by a smaller process change. Any of those can be reasonable when it follows the observed constraint. The point is to give the people asking for help a proposal that improves the service, rather than another team processing requests more quickly on their way to the same unresolved wait.