What you can tell about a company in a room that you cannot tell from a deck

A deck tells you what a founder has decided you should think. A room tells you how they behave when nobody is managing the impression. That is the only reason to go.

Here is what is actually worth watching.

Real signal

The unrehearsed explanation. Listen while they describe the company to somebody who cannot invest. That version is shorter, plainer and usually more accurate. When the two versions differ a lot, distrust the rehearsed one.

Specificity under follow-up. Ask one question, then a second and harder one about the same thing. Founders close to the problem get more precise as you press. Founders who are not get vaguer and start quoting the market size.

What they know that you would not expect: a detail about their supply chain, their churn cohort, their customers' internal politics. Depth in an unglamorous corner tracks time spent on the problem better than anything else you can observe.

How they talk about their team. Listen for named people, specific contributions and credit given without prompting. Founders who describe a team as a resource have usually not built one.

Being disagreed with. Watch whether they get curious or defensive. Founders cannot fake this in the moment, and it does not appear on paper at all.

Who they came with, and how they treat them, especially anybody junior to them.

Things that feel like signal and are not

Charisma in a room. It predicts fundraising ability and little else. Plenty of good founders are uncomfortable at events, and being good at a party says almost nothing about being good at building.

Pace of speech and confidence. Both are personality traits, and both get mistaken for conviction.

Name-dropping. Knowing who somebody is differs from being known by them. Watch whether those people greet the founder across the room.

A polished one-liner. It tells you they have pitched a lot, and nothing about the company.

Busyness. Companies that are working and companies that are thrashing both say "we are slammed", in those words.

Three questions worth asking

"What has changed in your thinking in the last six months?" Founders learning fast have an answer immediately. Founders reciting a deck do not.

"What is the part of this that is not working?" The quality of the answer matters more than the content. A precise, unflattering, unprompted answer is a very good sign. "Nothing really" is not.

"Who else is doing this well?" Founders who know their market name three companies and explain the differences. Founders who say "nobody, really" have usually not looked.

Writing it down

Four lines, the same evening, in the same place every time: what they are building, one thing they said that you would not have guessed, your read on them, and what would have to be true for this to be interesting in a year.

The last line is the one that pays. In eighteen months you will search for something, find this note, and know whether the thing that had to be true became true.