Notice what worked

Stated-Feedback Capture

Hear what users actually say, not just what they did.

Presents a feedback prompt — rating, survey, NPS, thumbs-up — to the people who interacted with your output, captures responses with provenance back to the specific output and respondent segment, aggregates across respondents, and routes the aggregated stated signal into your iteration loop. Fills the gap behavioural telemetry can't: what people say they experienced.

Shape

Operational dimensions

Human co-pilot

Person and system work side-by-side.

Event-triggered

Fires when an upstream condition occurs.

Low data gravity

Light state; replaceable any time.

Two-way integration

Reads from and writes to external systems.

Inputs

prompt template or survey instrument
target respondent population or trigger rule
provenance tie-back to specific output or interaction

Outputs

per-respondent stated response with provenance
aggregated stated metric (NPS, CSAT score, theme distribution)
routed signal to iteration loop or content team

Mechanism

Solicits and aggregates STATED feedback from people who interacted with the organisation's output (NPS, CSAT, surveys, thumbs-up/down, post-call ratings, user-research panels) and routes the aggregated stated signal back into the iteration loop.

Why this is a primitive

Cannot be decomposed — the elicit (form / prompt) → aggregate (across respondents) → route-to-learning-loop operation is one mechanism. It is structurally distinct from telemetry-refinement on TWO axes simultaneously: input modality (form / prompt vs event / log) and signal type (stated preference / reported experience vs observed behaviour). You cannot reconstruct stated-feedback from behavioural telemetry — they capture different facts (what the user SAID vs what the user DID), often disagree, and require different aggregation machinery (Likert / qualitative coding vs metric attribution).

Where it shows up

SaaS product — post-feature-use CSAT prompt capturing whether users found the new workflow intuitive, with responses tied back to the specific feature version

EdTech — post-lesson five-star rating plus open-text comment captured from learners, aggregated by lesson to surface which content needs rework

Customer support — post-call satisfaction survey triggered on case close, routing low-scoring calls to QA review queue

Content team — NPS pulse sent to newsletter subscribers after each issue, tracking satisfaction trend as an editorial health metric

Related primitives

Composes with

Telemetry-Driven Refinement Loop →Controlled Experimentation →Retrospective Capture →

Alternative to

Telemetry-Driven Refinement Loop →