Rotation

Perhaps the rule that is most important to the success of the Queue Master is that the duties must rotate across members of the team performing the work. Optimally this is done with a weekly cadence.

While at first many team members might find taking on a duty role to be disruptive to their own work, rotating the role at the natural cadence of a working week has a number of distinct benefits. The first is that it ensures the Queue Master maintains sufficient context with the work and team performing it. Active team members are exposed to the inner workings of the infrastructure, the software and services, and the people who manage them.

This exposure provides a level of insight that managers or anyone else not directly immersed in it lacks.

Rotating at a natural cadence also gives everyone a chance to step out of their day-to-day work and properly see what is occurring across the ecosystem. I cannot state strongly enough how eye-opening this can be. Most of us forget how easy it is to lose sight of the big picture when we are buried in our work. Stepping back from it to see what is going on across a typical work cycle not only helps us to see what is happening, but helps the team understand and resolve problems that damage team effectiveness.

Every time I have put in place the Queue Master rotation, I have had at least one of the team who was most skeptical of the idea come to me on day three or four of their first rotation exclaiming how they now understood why the role was so important, detailing how much crazy stuff they never realized was going on that needed to be fixed.

Rotating the role also helps everyone on the team start to see and really appreciate the knowledge and effort other team members are contributing. They will often see team members struggling with some problem or dysfunction that they can help with. They will also remember and appreciate when they are helped by another team member who takes on Queue Master duty. Together, this builds team unity and fosters further collaboration.

Rotation also brings value to the Queue Master role itself. The differing perspectives that each person brings to the role can uncover hidden problems and solutions. Rotation also reduces the likelihood that those in the role will get inured to the sorts of dysfunction that can so often damage team effectiveness. If nothing else, the shift in exposure to another perspective of the ecosystem is likely to spur people to question the status quo.

Rotation greatly reduces the chance of Queue Master specialization taking root in the team. The last thing any technical team needs is yet another specialization that reduces the flexibility of the larger team.

Entry Management

To be effective, the Queue Master must sit at the point where all non-incident work officially enters the team. This includes all work, from requests coming in from other teams to work internally generated by the team. The dual intent of this entry management is to ensure the clarity and priority of the work and to capture cross-organizational nuances that might otherwise be missed.

Entry management achieves a number of objectives. For one, it makes all work trackable and eliminates the chances of any assumptions cropping up that can be missed along the way. It also prevents people from misrouting tasks to any random person in the team, thus reducing interrupts and misprioritization of work. Another benefit of routing everything through the Queue Master is that they can act as a quick filter that helps limit garbage from entering the workflow. This reduces the chances of team members getting partial or incorrect information, or a misprioritized task hitting the team directly.

This does not mean that routing should be a heavyweight process, or that it is the Queue Master’s job to write up all the tasks themselves. Routing should be extremely light touch whenever possible. In some cases, particularly for internal team work, routing might be as simple as someone mentioning that they are adding a task to help clarify or add some important missing piece of work. As for writing tasks, the one requesting the work should write and submit the task whenever possible. The Queue Master can, if necessary, then follow up to add any missing information. This reduces any information loss.

All of this helps prevent the clogging up of the workflow. Most people might feel that their request is both important and obvious, even when it is not. This checking and sorting ensures that everything goes in the right place with enough information that it doesn’t get lost, misprioritized, or lead to unnecessary confusion. Critically urgent tasks will still always go into the Expedite column. The Queue Master can and should provide background and some queue reordering for anything else that might still be important or is a dependency to something that is yet not worthy of being expedited.

The best way to prevent unnecessary task rejection is for the team to proactively create and tune templates and examples of what a requester minimally needs to include in a ticket for it to be acceptable. This will go a long way in reducing both frustration and rework. Training and outreach should also be available to educate others across the organization. This will help stop the inevitable unhelpful tickets such as “The service is broken. Please fix,” “Need some software installed,” or the all-time favorite “Is the thingy backed up?”

Sorting and Dependency Discovery

As unplanned work tends to be a regular occurrence for service delivery teams, chances are good that work will come in that bypasses people like managers, project/program managers, and architects who can call out any hidden dependencies or ordering conflicts. To solve this, the Queue Master regularly goes through the queues in search of such problems to proactively help the team avoid any unnecessary blocks or rework that they might cause.

Having someone with intimate knowledge of the ecosystem do this has the added benefit of also catching any conflicts that might have been unknown or unclear to managers and architects. Sometimes this can be done in a lightweight way, such as simply flagging a conflict to the rest of the team to be aware of. In the case that the conflict is more serious, the Queue Master can send the task back to the requester for clarification or escalate to management for help.

To deal with all of this sorting, the default state of any work coming into the team is typically considered Unqualified, at least until the Queue Master has a chance to take a quick look at it to determine whether or not it is suitable for the board. Most of the time the Queue Master will determine the work item is fine and move it into the Ready column on the board. If, however, a task is just plain wrong or misdirected, the Queue Master can reject it and hand it back to the requester with an explanation.

Likewise, if a task is unclear, too big, or open-ended, the Queue Master has an opportunity to send the item back to the requester to clarify and improve the request before it has a chance to cause any problems with the team. One method for doing this is to alert the requester, then flag the task and place it in a Clarification queue in the workflow.

I have found that tracking both clarifications and rejections can help the team remain aware of what is going on while also helping them spot and understand if there are patterns or sources that are causing such problems that need to be addressed. This helps the team to focus only on the sorts of template, education, and process improvements that will make a difference in reducing problem occurrence over time. This improves their cycle time, but also reduces the frustration and workloads on the requester and Queue Master. It helps minimize any problems or misunderstandings that somehow make their way down to those performing the work, and does so without throwing in excessive processes for no real reason.

Improved entry management by the Queue Master also has the benefit of reducing the chances of work unevenly flowing across the team. Lean practitioners know that uneven flow creates waste. By ensuring that tasks enter in only through the Queue Master, team members avoid being put in a position where they have to choose between creating conflict with requesters and taking on too much work. This reduces team pressure.

Dark Matter Handling

One of the powerful benefits of insisting that all noncritical work must go through the Queue Master is that it becomes a single point to receive and handle all Dark Matter requests. This ensures that they are captured and tracked to aid in understanding their sources and occurrence.

Redirecting Dark Matter to one central place can be tough to do at first for both the organization and the team. However, doing so is worth all the grumbling and minor inconvenience. Not only is recording the tasks important for understanding the source of and sizing the demand for various activities, but offloading trivial tasks to one place reduces the number of interrupts that otherwise would pepper the team. This reduces errors and helps improve responsiveness and flow.

Maintaining Flow

The role of Queue Master does not end with work entering the workflow. In fact, many of the most valuable Queue Master contributions occur by keeping the focus on the bigger picture. One of these contributions is maintaining the flow of work.

For instance, not all tasks that reach the Ready column in the workflow board necessarily deserve the same treatment. While management, the product owner, or even individual engineers often do most of the prioritization work, there are some cases where the Queue Master may need to do some reordering. This is usually due to either sudden shifts in priorities or to ensure that one or more key prerequisites get completed to avoid workflow blockages. When a Queue Master does this, they will usually alert the team of the situation, either in the daily standup or via chat, to make sure that any priorities or dependencies are clear and understood.

If a task requires special handling because the work item requires either specific skills or specific context, the Queue Master can make sure that these handling requirements are clearly marked. This will aid in informing the team when the task is ready to be pulled in, and can help with tracking as it moves across the workflow. As a team matures and the Service Engineering Lead role (covered in Chapter 9, “Service Delivery Maturity and the Service Engineering Lead”) becomes better established, the team itself frequently becomes the biggest source of such tasks. They are usually related to delivery tasks created and marked by the SE Lead to help with tracking.

While everyone is responsible for the board, there are times when there are items that can languish in a column. The Queue Master actively monitors the board to spot them and remove any blocks or potential problems that might be preventing them from progressing. The Queue Master also looks for large build-ups of work in progress, especially those caused by unexpected problems that develop.

The Queue Master uses the daily standup, as well as the team’s retrospective, as a way to bring up issues the team members encounter. The Queue Master can ask team members for details, as well as help to remove the issues and restore flow. The Queue Master can also work with the team to understand the issues’ root causes and find ways to prevent them from happening again in the future. Together, these ensure that work does not languish unnecessarily and that team members become overwhelmed with work.

Pattern Recognition

One of the major advantages of having a Queue Master tasked with actively surveying the workflow from the entry point through to completion is that they will start to see patterns emerge that can go a long way toward helping the team improve.

Such patterns begin with the demand coming in itself. There is always subtle yet important interplay between the types of tasks coming in, those asking for them to be performed, and larger developments taking place across the ecosystem. This context is valuable, especially in environments where a significant segment of demand is not otherwise easy to plan for in advance. If you can understand the source of the demand, its frequency, and what might drive it, you can begin to predict when it might occur again in the future. It also gives insight into similar requests, and similar requesters, that helps you establish larger patterns of demand. This can aid in resource planning, as well as sizing investment return from standardizing and automating the task, or eliminating its root cause.

Similarly, you can also start to understand and track dependency patterns between tasks. If a particular task enters the flow, you know that there is a good possibility that those same dependent tasks might follow. Understanding these dependency patterns can help preempt additional requests and improve team responsiveness and understanding. This reduces workflow noise and provides useful automation targets. It also aids coordination.

Being able to uncover demand patterns also provides useful clues about what the requester is trying to accomplish with the request. This is important because not everyone shares or is clear about the outcome they are trying to attain. It is not uncommon for people to make requests that, due to solution bias or some other misunderstanding, will not achieve what they are after. The Queue Master is uniquely positioned to notice such discrepancies and flag them for further investigation.

Patterns also exist in the workflow itself. There might be tasks that take far longer than they should, require rework, or have a large number of handoffs. There are also times when the Queue Master might notice that certain tasks can only be done by one person or a small subset of the team, creating a dangerous bottleneck that indicates knowledge and awareness might not be flowing evenly across the team.

Each of these patterns should be noted by the Queue Master and discussed with the rest of the team through the various syncing and improvement mechanisms discussed in Chapter 14, “Cycles and Sync Points.” Queue Masters who are great at pattern matching can make a big difference for a team. As more people take on the role, people start to share and find patterns based on their experiences. This strengthens the collective mind, providing fertile ground for innovative solutions to challenging issues.

Office Hours

It is important to point out that, unlike a typical on-call rotation, the Queue Master generally is not a role that must be available 24 hours a day. Nothing the role does is so urgent that it must always be attended to. Even if there happen to be off-hours requests that are so critical that they need immediate attention, they can easily be routed to on-call.

For that reason, it is worthwhile to set the expectation that the Queue Master will be around and available at a set time period during the business day. Usually Queue Master work begins after the daily standup and ends at a reasonable set time each day that corresponds to the lull of task requests at the end of the workday.

Besides preventing burnout, establishing office hours also makes it obvious when work hits the team at unreasonable times. This doesn’t mean that there shouldn’t be a mechanism for extremely critical work that happens in the off-hours. In fact, if something so urgent comes in that addressing it cannot wait until the next day, it should be treated as a production incident. This does several things. It ensures that it will get immediate attention. It also ensures that the request is subsequently reviewed to see if there is a genuine need that has been missed that requires a change or even a new tool in order to address it.

Finally, and most importantly, raising an off-hours request as an incident can help requesters consider the urgency of their request. This allows for true problems to be accounted for and addressed in more sustainable ways.

Sync Point Management

Maintaining situational awareness, learning, and improvement all require there to be a certain amount of alignment of the team. There are lots of mechanisms that you can use to do this, from wikis and document repositories to chat rooms, standups, pairing, demos, meetings, and just working in the same location and chatting from time to time. Each has its advantages and disadvantages, which is why I have dedicated an entire chapter (Chapter 14, “Cycles and Sync Points”) to talk through patterns that have worked well in organizations I have worked. In them, the one common thread that I have found is that it is extremely helpful to have one person making sure that they happen. The best-placed person for that is the Queue Master.

The Queue Master will be exposed to the chaos at a high level, meaning that they should be better placed to see problems that others are living with but might be too close to see. By being the one in charge of the sync points, the Queue Master can help draw out the deeper details from others. This helps people get a better perspective and understanding of what is going on, helps build situational awareness, helps improve team collaboration, and even helps those in the heat of a situation to step back and look more objectively.

Other Work When on Duty

One challenge of a busy Queue Master cycle is that it might not give the person holding the role enough time to handle their normal day-to-day and project responsibilities. Like being on-call for production support, Queue Master responsibilities take precedence over just about everything else. This means that any other activities that the person has must either be transferred to someone else during the cycle or wait until that person has available capacity to continue.

This approach might seem extremely disruptive at first. However, as the Queue Master starts to make things visible enough that they can be worked on and improved, most can find a good balance between their rotation and their existing work.

One important thing that everyone should keep in mind is that the ability of the Queue Master to reduce the noise that otherwise would interrupt the team is of immense value. It means that the team members can complete their work without interruption, improving overall team flow throughput. Team members are also far more likely to share details about the work they are doing if they might need others to jump in to pick up the slack from time to time. This further builds team situational awareness and flexibility.

The cyclic rotation of the Queue Master role means the period is temporary and that activities can be planned to work around it. Over time as the workflow begins to settle, even this disruption will greatly lessen.