E Evidence Press

PRODUCTIVITY PROTOCOLS · COMPANY STARTER

Company starter kit

Choose the evidence stage before inviting people. Learn first; compare only when the design can support it.

Staged evaluation programme — not yet fielded

No company has yet run this programme with staff. The stages below are a prospective learning sequence, not evidence of adoption, productivity or impact.

Formative usability

Use with 1–5 consenting participants to find comprehension, operation, stopping and support problems.

No effect estimate

Feasibility

Use the frozen task bank and allocation records to rehearse recruitment, measurement, cost capture and retention.

No powered benefit claim

Controlled signal

Prepare a separate, justified and independently reviewed randomized parallel B–C design.

Not supplied by this rehearsal

Field follow-up

Observe use, burden, costs, errors and outcomes only after a separately governed deployment decision.

Association is not causation

Run it in this order

  1. Choose a routine, low-consequence task with a named owner and a result the team already knows how to judge.
  2. Record affected people, approved data location, retention, human approval, rollback and stop conditions before using an agent.
  3. Choose formative or feasibility. Do not promote either route into an adoption decision based on an estimated productivity effect.
  4. For feasibility, freeze the plan, task bank, configuration, prompts, rubric, allocation and analysis before work begins.
  5. Retain every planned work item, including missing, abandoned and technically failed items; record total human effort, tool cost, errors, burden and incidents separately.
  6. Review only the next-stage decision: continue one stage, revise and rerun, or stop.

Read before downloading