Live properties
These are online now. Every property is a separate project with its own owner, codebase and deployment. DatumHarbor observes them; it does not modify their code or product logic.
| Property | Domain | Question being tested | Status |
|---|---|---|---|
| Money Machine Intelligence | money.datumharbor.com | How real businesses create value and generate revenue | Live |
| Commercial Failure Intelligence | failure.datumharbor.com | Why commercial efforts fail, and what is reusable from that | Live |
| Problem Discovery Engine | problem.datumharbor.com | Which problems are commercially real rather than assumed | Live |
| Qianmen — Defensive Scam Intelligence | scam.datumharbor.com | How scam patterns can be recognised and defended against | Live |
| Enterprise AI Work Intelligence | aiwork.datumharbor.com | Where real enterprise AI deployment actually stalls | Live |
| Global Innovation Radar | radar.datumharbor.com | Which innovation signals are worth acting on | Live |
| 生意标本 (GSBI) | case.shengyibiaoben.cn | 真实小生意的赚钱机制是否可以复盘复用 | Live |
Formal per-experiment records — hypothesis, target user, commercial question, lifecycle state, decision and dated timeline — are maintained in the control platform and are not published on this site yet. Until a record has evidence behind it, this page shows only what is directly verifiable: that the property exists and is live.
What a registered experiment will contain
When records are published, each one will state:
The claim
What is being tested, for whom, which problem, and the commercial question it should answer.
Where it is
One lifecycle state, from idea through live testing to proven — or closed.
What was observed
Search, traffic, behaviour and commercial signals, each tied to the evidence behind it.
The decision
Continue, hold, or stop — with the reasoning and the date it was taken.
A stopped experiment is still a result
We expect to stop things. Stopping early is the point of testing cheaply. What must never happen is losing the reason: a stopped experiment stays in the dataset, and its evidence is never deleted.