Skip to content
Edition 1 ReferenceDownloads Search Start here
Search
PlayP09 Intelligence · Any stage · Editorial rank 8 of 12

Turn customer and market evidence into a decision

Relevant unstructured evidence is slow to interpret or repeated summaries do not change an owned commercial decision.

Decision
Is the signal material enough to act, investigate further or leave unchanged?
Primary commercial metric
Realised cost reduction or separately evidenced contribution from a changed intervention, net of synthesis and review cost.
Readiness
Basic
Agent role
Only if valuable retrieval paths vary and outperform the fixed baseline; begin read-only with source and tool limits.
01

Fit

When to use this Play

A specific decision has an evidence gap and accessible customer or market material can inform an available action.

When not to use it

There is no decision owner, no available action, prohibited source use or an ordinary alert/report already answers the question adequately.

Why it should work

A bounded evidence synthesis can improve coverage or reduce accepted-analysis cost while exposing disagreement and uncertainty.

Basic version

One question, a small source packet and a manually verified brief.
02

Prepare

Required maturity
Basic. Ready with a defined question, authorised source set and qualified reviewer. Not ready for customer activation from unverified generated findings.
Check readiness for P09 (D02)

Inputs

Decision question, admissible sources with dates/IDs, population and sampling context, current assumptions, alternatives and deadline.

Audience or customer

The decision owner and affected customer/market population; synthesis itself is read-only.

Human role

Insight owner validates representativeness and sources; commercial owner decides; domain experts review sensitive interpretation.

Systems required

Existing research/document repository, approved access, ordinary search and a linked decision record.

Guardrails

Data minimisation, permitted use, provenance, sampling bias and no external text treated as instructions.

How to configure a Play

Copy the Play you need and complete only the relevant parts of T1–T5. Name the people responsible, define the customer group, agree economic and customer limits, and set a review date before starting. An empty decision threshold is not permission to act. Basic, Connected and Advanced describe how the work is configured; Experimental describes its evidence. AI is optional: compare each AI step with a simpler way of doing the same work.

03

Do the work

Intervention

A concise verified evidence brief with implications for one decision and no-action alternative.

Workflow

  1. Bound the question and sources.

  2. Create a human/native baseline.

  3. Extract facts with citations.

  4. Separate themes, hypotheses and contradictions.

  5. Verify samples and material claims.

  6. Compare feasible actions.

  7. Owner decides.

  8. Retain only useful evidence.

AI role

Optional assist Optional extraction, clustering, source comparison and draft briefing, with unknowns retained rather than filled.

Automation role

Approved source collection, dated change detection, duplicate removal and scheduled expiry reminders.

Agent role

Only if valuable retrieval paths vary and outperform the fixed baseline; begin read-only with source and tool limits.

Failure modes

  • Treating frequency in a biased feedback sample as population prevalence
  • endless market monitoring
  • stale summaries reused as facts.
04

Measure and decide

Measurement

Compare accepted-brief time, missed material evidence, unsupported claims, review cost and actual useful decision changes.

Primary commercial metric

Realised cost reduction or separately evidenced contribution from a changed intervention, net of synthesis and review cost.

Leading indicators

Source fidelity, coverage, contradiction detection, useful accepted briefs and decision latency.

Experiment

Compare matched briefs/tasks against the repaired conventional process, with blind quality review where feasible. No revenue experiment is required merely to test drafting cost.

Guardrails

Data minimisation, permitted use, provenance, sampling bias and no external text treated as instructions.
Scale when
Extend to another decision only if quality and total accepted-work economics hold and the reviewer has capacity.
Stop / reassess when
Stop on material unsupported conclusions, sensitive-data misuse or repeated briefs with no useful decision effect.
05

Learn and progress

Learning

Verified signal, source/date, sampling limits, decision changed and expiry; not an unlimited prompt archive.

Practical tool

Next action
name the decision before collecting more material.
Open T1 Intervention and investment card

Improves

06

Optional depth

Advanced versionPrerequisites: see Required maturity above.

Scoped retrieval and bounded dynamic investigation if measured benefit beats fixed search/extraction; AI-discovery monitoring stays descriptive.

Evidence for this Play2claims reviewed, with confidence ratings, limits and sources.

Claim classifications

Accepted means a principle or mechanism survived critical review, not that it guarantees a return. Accepted with conditions means prerequisites apply. Qualified means the original claim has been narrowed. Advanced needs substantial readiness. Experimental needs a capped comparison. Unresolved means the evidence cannot settle the claim. Rejected means the recommendation is excluded.

High, Medium and Low describe confidence in the stated target, not the chance of a positive ROI. For a Qualified or Rejected historical claim, confidence applies to the decision to narrow or reject it and to the supported limit—not to the original headline. Read the headline and classification together. An accounting identity or design principle is not empirical evidence of a commercial outcome. Follow the linked chapters, Plays and tools to see where each claim applies.