It is your file, not a demo account
Bring one real file, even a redacted one. A twin built from your profit and loss behaves differently from one built from an invented company, and the differences are the interesting part.
Walkthrough
Forty minutes on a call, with your file open. A twin gets built while you watch, the question you brought gets run, and you get told where the answer is weak. If that is not useful you will be told that too, on the call, rather than afterwards by email.
One decision and one file is enough. You do not need a data room.
Setting expectations
Bring one real file, even a redacted one. A twin built from your profit and loss behaves differently from one built from an invented company, and the differences are the interesting part.
There are no slides. There is no pipeline stage to move you into. The call is somebody opening the software, building a twin from what you brought, and running your question while you watch.
The most useful thing you can do is open the ledger and disagree with things. Every number is editable, and watching the answer move when you replace a default with your own figure is the fastest way to judge whether the model is worth anything.
Forty minutes gets you shape, size and mechanism on one question. A result worth putting in front of other people needs your real numbers replacing the defaults, which is work you do afterwards.
If your question depends on something the model handles badly, for example a tender cycle rather than price driven demand, that gets said on the call. There is an entire page about the limits, written before anybody is asked for money, and the call is consistent with it.
This is new software built by one person. The call is with that person. That means no handoff to somebody who does not know how it works, and it also means there is no army behind it, which you should factor in.
Roughly. It bends to whatever you brought.
What you would actually do differently depending on the answer, and what would have to be true for you to change your mind. This determines everything that follows and it is the field the form asks for.
Your file goes in, the facts come out with their quotes, and the graph and the agents get built while you watch. You will see the readiness score and what is missing.
The scenario, the band, the events with the agent that caused them, and a sweep if the question has a lever with a range. Then the ledger, sorted by origin, so you can see what the answer is standing on.
You change the assumptions you think are wrong and the run goes again on the same seed. This is the part that tells you whether the software is useful or just confident.
Nothing on the walkthrough is hidden from a normal account. If you would rather not talk to anybody, start, upload a file, and the software will tell you what it found.