Start Here The shape of the methodology
One page. The whole system. Where you are, and what comes next.
Signal-Response Distance (SRD) measures where organisations lose value between a signal arriving and a response landing. It measures the delay, names the binding constraint, and tells you where AI changes performance and where it makes things worse. Everything on this site is one of five layers: the theory, the method, the diagnostic that applies it, the response layer that acts on it, and the evidence that tests whether it worked.
Each layer gives the next one its authority.
- 01
Theory
The Signal-Response Distance construct: what it claims, where it comes from, how it could be proved wrong.
- SRD Working Paper On request
abstract, lineage and version record public; full text on request
- SRD Working Paper On request
- 02
Method
How a practitioner applies the construct: the five-phase protocol, the working surfaces, the patterns that compound across applications.
- SRD Application Playbook Public
structural tier open; the operational tier (calibration, thresholds) stays in the application layer
- Engagement Canvases (v0.8) Public
six A3 working surfaces, free download, CC BY-ND
- Pattern Library Public
seeded; grows only from consequentiality-reviewed applications
- SRD Application Playbook Public
- 03
Application
The diagnostic that runs the method on a real organisation and decides, by verdict, where intervention belongs.
- SRD Value Diagnostic Public + engagement
the reading runs openly on this site; the full calibrated instrument runs inside engagements
- SRD Value Diagnostic Public + engagement
- 04
Execution
Acting on the verdict: governed response, measured against the diagnostic's own clock, never preceding it.
- Response layer Engagement
delivered inside engagements; the diagnostic decides where it goes
- Response layer Engagement
- 05
Evidence
Whether it worked: consequentiality follow-up at 6 and 12 months, claims that survive, get amended, or get retracted in public.
- Research agenda and open questions Public
the road to v1.0: a third materially different case, then the inter-analyst reproducibility study, then journal submission
- Research agenda and open questions Public
It is not a pipeline. It is a flywheel.
Evidence does not sit at the end; it feeds back to the start. Every consequentiality review amends the patterns, the patterns recalibrate the diagnostic, and claims the evidence disconfirms are amended in public rather than quietly retired. That closed loop, not any single layer, is the asset.
Structure public. Calibration private.
The open layers show the structure of the method: the construct, the protocol shape, the working surfaces, the patterns. The application layer (analyst calibration, scoring rules, thresholds, reproducibility protocol, evidence weighting) runs inside engagements and, on request, in front of named reviewers. The canvases are complete enough to run an uncalibrated application; the calibrated one is what an engagement buys.
The named commercial door is the AI Value Sprint: ten days, fixed fee, stop/fix/scale verdicts on one AI portfolio before its funding gate. Delivered by the practice at ajolivier.com.