Meeting Effectiveness Survey Questions That Actually Show Something

A meeting effectiveness survey asks the people who were in a meeting whether it achieved what it was called for. Good questions cover eight things: purpose, preparation, participation, decisions, duration, attendees, follow-up and overall worth. Bad questions ask about satisfaction, and satisfaction is the one thing that does not predict whether the meeting was useful.

That distinction is not academic. In the classic HBR survey on the subject, 71 % of senior managers said meetings are unproductive and inefficient, most often because a meeting meant for a decision is used for status sharing instead. A satisfaction question cannot see that. A decision question can.

Below are the eight questions, what a low score on each actually means, and the three questions that are worse than asking nothing.

Rate each on a five-point scale. The right-hand column is the part most question lists leave out: the same low score means a different problem depending on which question produced it, and the fix differs accordingly.

Read the spread before the average. A 3.5 built from fives and twos is a different meeting from a 3.5 built from threes and fours, and only the first one has a participation problem. That split usually means some people in the room are having a good meeting and others are attending one.

Then read pairs rather than single numbers. Decisions against follow-up tells you whether you decide well and forget quickly. Participation against overall worth tells you whether the people who stayed quiet also found it pointless. If participation is consistently low, the underlying question is not about meetings at all: it is whether people expect speaking up to be worth the risk, which is what psychological safety measures and what Google identified as the top factor in team effectiveness.

And watch the calendar, not only the scores. HBR's work on managers holding too many meetings makes the point that meeting load is a decision someone made, usually years ago, and never revisited. A team mood check or a running pulse survey will show the load effect before any individual meeting score does.

Eight ratings tell you that something is wrong and roughly where. They do not tell you how the group actually talks: who picks up whose idea, where a decision was made and then quietly reopened, which objection was raised once and never again.

That needs the words, not the scores. If a transcript of the meeting already exists, reading it against a framework is the next step, and it adds no new recording and no new consent question. We wrote up which tools produce those transcripts, where they store and process them, and the German and Austrian legal steps involved, in AI meeting notes and GDPR. For the reading itself, the C&C meeting analysis examines an uploaded transcript for atmosphere, decisions and feedback loops. Where the issue is the leadership circle rather than the meeting format, the manager effectiveness guide is the better next read.