A simulated public, on the record
Meet the room before you walk into it.
Glasshouse shows you how people will react to a decision before you make it. Read who pushes back, who comes round and why, every claim traced to its source.
How it works · Step 1
Tell it the decision.
Describe the decision in a chat and attach the documents behind it. Name the voices you must hear from; they are always in the room.
We want to ship automatic shift swapping next quarter. Will schedulers trust it, and does it move renewals?
Step 2
Read what it will rehearse.
Glasshouse drafts the scenario: the question the room will answer, and what your evidence can and cannot support. Nothing runs until you approve it.
Will schedulers trust automatic shift swapping, and does it move renewals?
Step 3
See who cares.
Glasshouse keeps a Library of stakeholder profiles and polls them on your scenario: who would care, and how much. The most interested become the cast: synthetic personas with a job, a stake and a temper. Nobody in the room is real, and everything they say is on the record.
Step 4
Watch it argue.
The cast reacts over timed rounds, the way a public does online. A rehearsal spans one to six simulated platforms at once. Personas post, reply and change their minds. Narratives form and turn while you watch.
Automatic swaps only work if the app knows who is qualified for the shift. Ours does not, yet.
It does if you let it read the certification field. That is the whole feature.
Six months of swap requests handled by hand. If this works I renew without reading the invoice.
Before opening night
The world reacts once.
You can rehearse twice.
Step 5
Read what the room decided.
The rehearsal ends in a reading: a recommendation, its confidence, and the numbers behind it.
Recommended: build it, and ship the qualification check first.
No black boxes
Every claim keeps its receipt.
Behind the reading sits a report you can hand to a board, with a footnote on every claim: who said it, and in which round. No source, no claim.
Trust in the swap turns on the qualification check, not on the automation itself. Tobias Brandt R9
Automatic swaps only work if the app knows who is qualified for the shift. Ours does not, yet.
Know how much to trust it.
Every report carries Diagnostics, a plain reading of the run itself: did the narratives settle, was the cast balanced, how deep is the evidence. A decision aid, not a crystal ball. If the room was still arguing when the budget ran out, Run longer adds rounds to the same cast and past, price shown first.
Same cast, same past, new rounds. Price shown first.
Run longerStep 6
Ask the room why.
Pick one voice out of the crowd and ask why it said that, then what would change its mind. It answers in character, and every answer cites the run.
A Pro control.
Why did you post that?
What would change your mind?
Pro
Take the controls.
Set up the room yourself.
How many take part, which platforms, how deeply. The price follows the room.
Ask what if.
Change one thing: the price, the timing, who speaks first. Run it again and read the two futures side by side.
Fork from any round.
Step back, introduce news or a statement of your own, and run a second future. The original stays.
Read all of Diagnostics.
Every quality reading behind the report, fully audited by Glasshouse agents.
Plans
Three ways in.
For one person rehearsing one decision.
- Tuned by us, nothing to configure
- The report, every claim footnoted
- Diagnostics
- Run longer
For the person who runs the room.
- Everything in Basic
- Set up the room yourself
- Fork, introduce news, compare futures
- Interview the cast
- Advanced Insights and Diagnostics
For a team, with a named contact.
- Everything in Pro
- Much higher credit limits
- Terms for teams
Who it serves
If people will react to it, you can rehearse it.
A product team asks whether the feature it is about to build will be adopted, and whether it earns anything. A company asks how four separate products land as one bundle. A team asks who would buy an MVP that does not exist yet. An executive stress-tests an AI strategy. Communications and policy teams rehearse an announcement, a draft law, a consultation.
We work this way ourselves: major Glasshouse feature decisions are rehearsed against a simulated public, for reception and for revenue, before we build.
Early access
Rehearse the next decision.
Meet the room before you walk into it.