Work Entry: Jumping the Queue

Over the past few months, I have been in traffic jams on the highway several times when traveling to our weekly hike.  In more than one instance someone has decided to pull over and drive on the berm.  In more than a few cases the outcome of this technique for getting things done ends poorly. Despite the unpredictable outcome, jumping the queue is practiced by many in traffic and even more when funneling work to teams. The consequences in information technology are far more predictable than driving, and they are ALWAYS bad.

Jumping the queue happens because:

  1. There isn’t an effective prioritization approach,
  2. Starting is more important than finishing,
  3. Venerating getting stuff done without regard to the “how”,
  4. The impact is not publicly shared, and
  5. The end of the world is nigh if we don’t act now.

Without an approach for prioritizing work, it is easy to make a case that something on the backlog is more important than something you are currently working on. This is a problem at every level of an organization, from strategic initiatives to user stories and tasks, when there isn’t a clear goal and a pragmatic and repeatable approach to prioritizing work. Approaches such as weighted shortest job first, cost of delay, and value ranking are useful approaches for prioritization. An organization should have a common approach for prioritizing each level of its portfolio because failing that everyone will have their own way of doing things, and they will be apt to change whenever someone speaks loudly.

Starting every request is easier than telling people they have to wait until there’s capacity. In scenarios where jumping the queue is ok, waiting your turn in line is a bad idea because you may never get to the head of the line. I have traveled in several countries where queuing to get on a train or plane is not a thing.  The first few times I ended up sitting on my carry-on, now I jockey for position with the best. I do better now, but those behind me lose out.

If you have ever seen the TV show MASH, you will remember the character, Radar O’Reilly. One of Radar’s claims to fame was that he could get anything.  He knew all the tricks. The same pattern of behavior happens in many organizations, the person that gets their project or enhancement delivered is a winner while those that wait for their turns are often considered wieners. Jumping the queue generates higher work-in-process, slower cycle times, and higher technical debt which reduces the value any team delivers to the organization.

The impact of jumping the queue is rarely measured, let alone advertised when known. What gets talked about are the successes, not the cost. When a resource or behavior is thought to be free or doesn’t have negative consequences it will be overused until the pain is overwhelming. This is a form of the tragedy of the commons. One of the costs is that work that has higher value will be delayed or abandoned. Organizations take real money off both the top and bottom lines by letting work jump the queue.

Sometimes jumping the queue makes total sense. I was recently talking to a friend when he got pinged that the corporate website had taken a hit and was down. He had to shelve plans for the day. I was working at a firm when their data center flooded; plans changed on the spot. When the end of the world is nigh, plans are going to change. I included this reason for fairness, but it is uncommon. Always remember that there is a difference between urgent and important. Make sure the work that gets into any workflow favors the important rather than the urgent.

Jumping the queues is one way even mature work entry processes get crushed which then cascades into other problems. 

Tactics to stop or limit queue jumping

Jumping the queue is a time-honored approach for getting your work addressed before planned work. This behavior happens due to a variety of reasons.  There are only two valid reasons (albeit one is tenuous).

  • The business environment has changed. This is the tenuous reason. It is tenuous because a better response to changing business conditions is to step back and reprioritize rather than to willy nilly jump the queues. Long periods between reprioritizing and planning can lead people to decide that the most effective means of addressing their perception that their work has become more important is to change the planning horizon. The solution is to break work into smaller components and reprioritize more often.
  • The end of the world is nigh if we don’t act now. There are times when things have to change. The response to Covid is an example where work had to jump the queue. This did not negate the need to then replan whole portfolios of work. These scenarios occur but are very rare. What is more common is poor leadership and/or failing businesses trying to hit the magical 5 run home run (not possible).  I have worked in my troubled companies over the years. In the end, each stumbled around looking for the magical product or service that would restore them to health (almost always at the expense of the less glamorous approach of reducing bloat and improving service).

Almost every reason for the perceived need for work to jump the queue is a failure in discipline. Someone believes their needs are more important than the larger organization or that they know more than those who made the decision.  Suggested solutions include:

  • Create and ENFORCE a policy that denies queue jumping. In Tame Your Workflow, Steve Tendon makes the case for a policy forbidding queue jumping, showing the impact of flow disruptions on all forms of work.  Just having a policy is worthless if it is not enforced and the enforcement has teeth. Saying something is wrong and then celebrating the rare success of the action creates a deeper moral hazard.
  • Break larger pieces of work into smaller or more independent pieces. If disruptions in the queue become necessary the probability of there being a natural breaking point in the near term will be maximized. When this step is done alone the goal is to minimize the impact.
  • Reprioritize more often at all levels of the work portfolio. Stay abreast of strategic (technical and market) changes and use that knowledge to guide the work that is being done. This step will cut the risk of uncertainty. Do this step in conjunction with breaking work into smaller pieces. 
  • Involve the right cross-section of stakeholders in prioritization. Having the right people will help to negate the impact of biases that might reduce the effectiveness of work entry and prioritization.
  • Have a transparent process for prioritizing.  Jack Welch said, “Trust happens when leaders are transparent.” Knowing how and why decisions are made makes it easier for decisions to be accepted and reduces the pressure to subvert the process.
  • Measure! Track the state of the portfolio and the value initiatives are delivering.  Channel funding and people to what is important and keep an eye on your decisions in case things change.

The best solution is to do all six solutions together.  I would suggest that five of the six can make a positive impact even done alone.  The only one I would avoid doing alone is the policy suggestion.  A policy without a process and transparency in the application will destroy trust. Rules that are not enforced lead to anarchy. Unlike the drivers that make me crazy by driving in the berm, the consequences of doing the wrong work or the right work at the wrong time can have huge organizational consequences. Leaders do not need to accept queue jumping if they make few changes to the work entry process and then pay attention to the results.