And 3 reasons why we could avoid a project disaster and frustrations of our project teams and clients came to my mind.
- We promise what's possible - small clients do understand that limitless features are misleading. They think in business results and need-to-have, not nice-to-have, features.
- We share worries from the beginning - small clients recognize our worries, say, about complicated models, spurious precision, correlation traps, … the toxicity of certain instruments. They want to deal with instruments they can manage.
- We polish based on co-evolution - requirements are dynamic and at small clients we talk to the decision makers. They don't hide problems. Take the Black vs Bachelier problem. It was discovered by the senior financial engineer of a regional bank, running tests with a widely used valuation system. The Bachelier solution is now part of the UnRisk FACTORY version 5.
Be clear on each other what is on the table. There is risk enough - no need to add operational risk from too much promises, information hiding and lack of polishing.