If you are discussing a topic with colleagues, it’s almost always better to have a specific conversation instead of a general conversation.
|General Conversation||Specific Conversation|
|“We need to figure out how to scrub all open bugs.” Followed by a lot of non-specific debate…||“There are 42 open bugs. 42 bugs fit on one screen in Excel. Lets look at them all right now and see if there’s a pattern.”|
|“Customers are angry. We need to put a plan together to make them happy.”||“Our biggest dealers in the NE region are angry about a bug in product X. There are four such customers, and the contact name for each are listed here. Let’s get a plan together for each.”|
|“Sally is hard to work with.”||“Situation: I sent Sally an urgent email last Tuesday on topic X. |
Behavior: Sally didn’t reply until today saying she couldn’t fulfill the request.
Impact: The project is now delayed and I don’t trust Sally to be responsive.”
I think you get the point. If not, add a comment!
Tools I use to try to get conversations out of the general, to the specific:
- Ask: “How many entities are we talking about here?” If it’s less than what can fit on a page or a screen DO NOT HAVE A GENERAL CONVERSATION. Dive in to the data right then and there.
- If the number of entities in the problem is truly large (hundreds) then suggest a taxonomy that will enable the group to bucketize. In a conversation the other day we were talking about employees in my org. That’s hundreds of people. Someone suggested “how many are managers, how many are engineers, and how many are other?” That broke the problem down so we could get specific.
- If you can bucketize the entities, prioritize the buckets. Very frequently you’ll find that you can completely ignore two out of three buckets, enabling a specific conversation about a small number of entities.
- Ask: “What’s a real example of this issue that exemplifies the issue?”
- When dealing with people, use the Situation-Behavior-Impact (SBI) model for feedback. See https://www.radicalcandor.com/blog/give-humble-feedback/ and this tweet-storm.