Escalations, whilst seemingly useful for creating a timely response, are actually a sign that there may be some underlying problem with the system of work. This can lead to a reinforcing loop of behaviour that will continually create longer wait times for your customers. You might not notice this at first, but some organisations get into a position where you have to escalate just to get anything done.
Escalations occur where teams run into issues and they need to bring it to the attention of various levels of management to get something done. There may be an urgent need that the service providing group in your organisation is not aware of or not aligned to.
Lets take an example where a team are running at their current capacity, as per normal, and a new request comes along. Although it may be more urgent, this team is already at it’s WiP limit and can’t take on anything more. They may not have different classes of service, or if they do, often these class of service items have already reached their WiP limit. An escalation occurs from the new requester. The team are asked to ignore their WiP limit “just this once” and fast track this item.
The result of this is that the team now have more WiP in their system. The lead times for every other work item are going to start to increase as the team fast track the escalated item. Lead times are getting longer and other teams waiting on work start to see their service level expectations slide past. So what do they do – they escalate. Now there are multiple escalations going on for the one team. Hopefully, sanity might prevail, but in many cases, this often leads to context switching by the team as they now try to service the other requests.
All of a sudden the team are now starting to struggle with all of the escalations and new requests. So, in an effort to try and get things under control, their manager attempts to reduce requests coming into the system. However, requesting groups also have obligations to meet, so guess what, they escalate.
What’s happening here is a self re-enforcing loop that see’s the level of service spiral out of control. Here’s a small diagram depicting what’s going on:
Urgent work, if handled correctly with WiP limits and classes of services can be manageable. When you get escalations for urgent work which ignore WiP constraints, you will likely fall into this kind of problem.
This is one of the things I learned from my Okaloa Flowlab training – if you’ve ever played the game you’ll get these kinds of insights and more (it’s a great experience). I’ll be including it in my upcoming Kanban System Design courses for those who want to experience it.
Try and break the loop by changing expectations to avoid further escalations, preventing urgent work coming into the system, or re-examining your capacity and policies around urgent work. They may also be hiding a larger problem in the organisation – around KPIs or lack of alignment more generally. Once you get the immediate problem under control, you may need to consider tackling the larger underlying issue.