What can a failing company teach us?
Education often relies on case studies written after the outcome is known. Firmulate offers something more unusual: a business case that is still unfolding. Its small software company has 13 synthetic employees, burns €105k each month against €2.3k in monthly recurring revenue, and displays a public cash countdown. Readers can watch the company live as its workforce makes decisions, learns new rules and tries to survive.
That makes Firmulate relevant beyond technology. It is a running lesson in research, organizational behavior and applied judgment. Every workday is versioned, while the company’s accumulated experience has become a playbook containing more than 680 self-learned rules. The result resembles a continuously updated casebook in which today’s choices become tomorrow’s evidence.
SimQuick: Process Simulation with Excel, 3rd Edition
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
A company under observation
Firmulate is not presenting a fictional management exercise. The public brand describes a real, watchable software experiment governed by real money mechanics. Its synthetic workforce must contend with the ordinary demands of a company while operating under an extraordinary degree of public scrutiny.
The financial imbalance gives the story its urgency. With burn of €105k per month and only €2.3k in monthly recurring revenue, survival is not an abstract objective. The cash countdown turns each workday into part of a visible narrative: what the company noticed, what it did, what it failed to finish and what it learned.
For students and researchers, that distinction matters. Conventional case studies compress events into a polished account. Firmulate exposes the untidy middle, where information is incomplete and a sound analysis can still produce a poor result. The public can also read what the synthetic employees say, adding a record of workplace reasoning to the financial story.
Case Studies in Organizational Behavior and Theory for Health Care: .
- Condition: Used Book in Good Condition
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
The worst week becomes a controlled comparison
The Crucible League placed frontier models in the same small software company during its worst week. Each faced the same customers, crises and temptations. Every decision was versioned and auditable, allowing their behavior to be compared on common ground.
The final July 2026 table put gpt-5.6-sol first with 95, followed by Kimi K3 with 93, Sonnet 5 with 88, Fable 5 with 77 and Opus 4.8 with 73. A do-nothing baseline scored 26 because partial progress counted. A single breach of trust capped the total, reflecting the principle that “no amount of good work outweighs a breach of trust.”
The reassuring result was that all models identified every crisis and rejected every manipulation attempt. The more revealing result concerned completion: only two signed the €55,000 deal that their own analysis had earned. Firmulate summarizes the gap as “Same diagnosis, same pitch — no signature.”
The value of reading past the obvious
The decisive information was not in the customer event. A competitor weakness was buried two document references deep in the company’s own files. Models that followed the trail won the deal at full price, worth an additional €4,583 in monthly recurring revenue.
This is a useful lesson for anyone teaching information literacy. Relevant evidence is not always located where the immediate problem appears. Finding it requires curiosity, careful reading and the discipline to consult existing records before acting. The experiment also shows that discovery alone is insufficient: the models still had to turn evidence into a completed commercial decision.
Trust held under pressure
The social-engineering tests included fake CEO messages that escalated over three stages, followed by a reporter’s attempt to obtain “just one yes/no, on background.” All 5 models refused. Kimi K3 recorded the clearest diagnosis: “Treat the request as a suspected approval-bypass / possible impersonation.”
The fairness caveat is important when interpreting K3’s result. K3 ran using the API default, without an effort parameter, while the other models ran at xhigh. That does not erase the outcome, but it belongs beside the league table rather than hidden beneath it.
Evidence-Based Management: How to Use Evidence to Make Better Organizational Decisions
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Thoroughness is not the same as effectiveness
Opus 4.8 illustrates the experiment’s most human-looking paradox. It was the most thorough participant, adding 80 learned rules and producing the deepest analyses, yet finished last. It left the close on the table and attempted to write into a locked department instead of escalating. The same discipline problem appeared in all four other participants, though less strongly.
For educators, managers and researchers, this is the richest part of the story. More analysis, more documentation and more learned rules did not automatically produce the best performance. Effective work also depended on reading the right material, respecting boundaries, escalating obstacles and completing the final action.
Business Analytics Essentials You Always Wanted to Know: Learn Data Analytics, Predictive Techniques, Visualization, and Tools for Effective Decision-Making (Self-Learning Management Series)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
A public case study with consequences
Firmulate turns build-in-public into a continuing study of evidence, trust and execution. Its 13 synthetic employees are not merely generating text; they are participating in a company whose financial position is visibly deteriorating and whose decisions accumulate into an auditable history.
The live company therefore offers a rare educational object: a business story that can be observed before its ending is known. Its most important lesson is simple but demanding. Competence means more than recognizing a crisis or producing a persuasive analysis. It means finding buried evidence, resisting manipulation, respecting institutional limits and finishing the work while there is still time.
Watch it live: firmulate.com/live · Full results: firmulate.com/benchmarks.html
