As part of a plan for the day, I pull together a list of meetings that will happen today and put an estimate on the meeting based on the scheduled time for that meeting.

As I've watch myself work these I've found that this approach grossly underestimates the impact of the meeting on my day for a number of reasons:

• Meetings typically require preparation. Unless the meeting is one that is on a cadence, meetings seem to require about the same amount of preparation as the actual length of the meeting. Even meetings on a regular cadence often need some level of preparation work (for example, Sprint Review requires that I put some words together to describe what happened during the Sprint).
• Meetings typically have follow-up actions. Unless the meeting has been mainly about communication of ideas, there will be some level of actions and recordings required from the result of the meeting. Even meetings on a regular cadence have these (for example, weekly Penguins meetings has specific actions for me, and a requirement to record the results in some place.)

My plan for today therefore typically includes:

• The actual meeting. The meeting itself is time-boxed. If you have taken the time to get others involved in the meeting, there must be some level of importance. The work is classified as (I)nvest so as you get to the end of the time-box you should ask “what value will we get out of extending this meeting?”
• A “meeting preparation” task which serves both as a reminder that something is coming up, and as a placeholder to actually do something. The typical work is administrivia or low level research to get a feel for what you think about the issue (without necessarily forming hard conclusions) and so the work is classified (O)ptimize. Sometimes a lot of research is required to prepare for the meeting, in which case you might look at additional tasks that are probably classified as (I)nvest.
• A “meeting followup” task to address actions, take care of any recording. Typically this is the administrivia as well as dealing with the actions means new user stories and so the work is classified (O)ptimize.

The impact is clearly larger than the meeting itself. For a 1 hour (~2 iterations) meeting, we typically have 1 hour (~2 iterations) of preparation work and 30 mins (~1 iteration) of administrative follow up work. On a fully booked 6 hour (12 iteration day) you are looking at 5 iterations for any meeting in the day,or ~40% impact on the day. Huge! The impact is probably more than that in that the meeting often represents a context switch and often (say 50% of the time on average) is scheduled so that I do not get a complete iteration to do work (ie interrupted, or more likely I simply don't bother to start an iteration.)

If this level of work is required (and I think it is in order to make sure that meetings I call are effective) then I need to question if I call a lot of meetings each and every day, am I really being effective?