Kanban on engineering dealing with customer services

Question

Let's say your team decides to go with Kanban, since the evolutive, adaptive and corrective work is under constant discovery, and planning for sprints is therefore not possible.

Along with that, you have a customer services team by the phone hearing the user's concerns, and filing issues in a tracking tool to our specialists team once they feel they received enough info.

What would be the best course of action to handle the fixed amount of work per state column if to make sure some issue is not happening again, the team depends on the customer's feedback (which sometimes takes ages to receive)?

That is either if customer services does not provide the team enough info to start an issue or feedback from the customer to close it.

One of my ideas was to keep a column for "awaiting client feedback", which should have a fixed size, and if the client does not answer in a timely fashion, the oldest of the items there gets closed as client unresponsive after a fixed timespan. But even then, we might have a bottleneck that cannot be resolved unless we remove the self-imposed limitation, right? ... Like considering the issues on that column as items out of your bucket. But wouldn't that break with one of the best added values of the framework?

Answer 1

My answer is there isn't a single answer so Experiment.

Don't forget the Kaizen part of Kanban. You should continuously be looking at what is working and not working and then tweaking the system with experiments. Change one thing at a time, so you see the effects.

One experiment could be to break your "Task" into two parts. The first task is the work up sending it off for customer feedback, the second is when the customer has given feedback. I recommend this because you have no control over the customer feedback. You don't want to limit work on a new task because you're waiting for feedback that may never come.