How it works
From your history to the dates ahead
Each question you bring us becomes a Forward Timing Report covering up to four months.
You bring the question
A launch, a campaign, an AI rollout, a supplier contract, a hiring drive, a negotiation, or any decision where timing matters.
You share your history
The dates of related past events and how each one turned out. A mutual non disclosure agreement is signed before any information is exchanged.
We run the analysis
Your dates are grouped by outcome and analyzed for the patterns in their timing.
We seal the prediction
The report is recorded with a public timestamp before you see it. A digital fingerprint proves it cannot be changed afterward.
You set the standard
You tell us what you will count as an accurate prediction. Only then do we release the report, always before the period it covers.
We review the results together
Each month we meet to go over what happened, score each prediction against your standard, and hear your feedback.
What you receive
- A summary of the most favourable dates, each with a plain explanation
- A timeline showing how your past events project forward
- A calendar marking every date on our favourability scale
- A detailed list of the dates that stand out, and why
Convergence
When you give us more than one history, for example past advertising campaigns and past sales results, we look for dates where both histories point the same way. Those convergence dates carry the most weight in our reports.
The favourability scale
Each date is ranked against every other date projected from your own history. Favourability is relative: it describes your company, your question and your period, not an absolute level.
| Increased favourability | The strongest dates in your run, the top 5% of all projected dates. The best windows to act. |
| Mild favourability | The next 15%. Good windows, with less behind them. |
| Neutral | Projected, but not standing out from the rest. |
| Reduced favourability | Dates that echo your setbacks rather than your successes. Windows to avoid or prepare for. |
| Mixed signals | Dates that echo both your successes and your setbacks. We tell you so plainly. |
Our commitment to proof
Anyone can claim to see the future. We prefer to show our work.
Sealed before you see it
Every prediction carries a fingerprint recorded on OpenTimestamps, a free public service anchored in the Bitcoin blockchain. You can check at opentimestamps.org that the report you receive is the one we sealed, and when.
Your standard, not ours
You decide what counts as accurate, before you see the dates. We never set the bar for ourselves.
Scored against chance
Results are compared with how often random dates would have met the same standard. A result only counts as meaningful if it clearly beats chance.
Misses are reported
Every prediction is scored, including the ones that do not land.