{
  "source": "Actual captured editor DOM for the fictional saved checklist",
  "html": "<h2>Review an Onboarding Pilot Before Extending It</h2>\n\n<h2>Introduction and Checklist Purpose</h2>\n<p>Use this checklist to decide whether a limited onboarding pilot needs more review before you recommend an extension. Operations managers and decision-makers should work through each item in order, using the supplied pilot record as their starting point.</p>\n<p>The decision is whether to extend a pilot-not whether to authorize a company-wide rollout. Treat the results as observations about the participants and period studied, not proof that the checklist caused a change. Before you begin, gather the pilot record and identify who will make the extension decision.</p>\n\n<h2>Pre-Task Preparation</h2>\n<ul>\n<li>☐ <strong>[Critical] Confirm the decision scope.</strong> Record that the proposal concerns a limited pilot extension, not company-wide implementation. The supplied record does not authorize a company-wide rollout.</li>\n<li>☐ Identify the operational question decision-makers need answered before choosing whether to extend.</li>\n<li>☐ Record what evidence would change the decision. If the current record is insufficient for the decision and no specific question can be answered by further observation, document the uncertainty.</li>\n<li>☐ Review the supplied evidence pack using <a href=\"/guide/quill\">Review the source pack before extending the pilot</a>.</li>\n<li>☐ If preparing an editorial summary, compare it with the content brief using <a href=\"/tools/brief-compliance-check\">Check a draft against its brief</a>.</li>\n</ul>\n\n<h2>Main Checklist Items</h2>\n\n<h3>Phase 1: Verify the Pilot Record</h3>\n<ul>\n<li>☐ <strong>[Critical] Confirm the cohort and period.</strong> PILOT-01 describes 24 volunteers observed over four weeks.</li>\n<li>☐ Verify that all 24 participants received the new onboarding checklist.</li>\n<li>☐ Record that the pilot had no comparison group. Do not describe it as a controlled comparison.</li>\n<li>☐ Keep conclusions limited to what happened for this volunteer cohort during the four-week period.</li>\n</ul>\n\n<h3>Phase 2: Check the Reported Results</h3>\n<ul>\n<li>☐ Confirm that PILOT-02 reports 30 support requests in week one and 18 in week four from the same cohort.</li>\n<li>☐ Label both figures as counts of requests, not counts of people.</li>\n<li>☐ Check the arithmetic: 30 requests − 18 requests = 12 fewer requests.</li>\n<li>☐ Check the relative calculation: (12 requests ÷ 30 requests) × 100 = a 40% decrease from the week-one count.</li>\n<li>☐ Describe the 12-request and 40% changes as descriptive results for those weeks, not as effects caused by the checklist.</li>\n</ul>\n\n<h3>Phase 3: Identify Evidence Gaps</h3>\n<ul>\n<li>☐ <strong>[Critical] List every unmeasured area identified in PILOT-03:</strong> time-to-productivity, completion quality, costs, non-volunteer outcomes, seasonality, and results after week four.</li>\n<li>☐ State that exposure and task mix were uncontrolled. Do not imply that either remained constant.</li>\n<li>☐ Record that the pilot design cannot show what would have happened without the checklist.</li>\n<li>☐ Keep outcomes for non-volunteers unresolved; the volunteer results do not establish whether the same pattern applies to them.</li>\n</ul>\n\n<h2>Critical Items to Double-Check</h2>\n<div class=\"callout warning\"><p><strong>Warning:</strong> Do not present the 40% decrease as a proven reduction caused by the checklist. Do not generalize results from 24 volunteers to non-volunteers or a company-wide rollout.</p></div>\n<ul>\n<li>☐ Check that the summary separates observed request counts from explanations for why they changed.</li>\n<li>☐ Verify that it does not claim the pilot established faster onboarding, improved work quality, lower costs, or lasting results. Those outcomes remain unmeasured.</li>\n<li>☐ Confirm that possible explanations for the request-count change remain unresolved, rather than appearing as findings.</li>\n<li>☐ Keep scope, period, units, design limits, and qualifications next to the claims they describe.</li>\n</ul>\n\n<h2>Common Items Often Missed</h2>\n<ul>\n<li>☐ Check that the counts refer to support requests, not individual employees.</li>\n<li>☐ Include the uncontrolled exposure and task mix when describing the week-one and week-four comparison.</li>\n<li>☐ State that no comparison group was used; repeating the same observation alone would not address that limitation.</li>\n<li>☐ Keep the absence of company-wide authorization visible in the decision record.</li>\n<li>☐ Avoid treating unmeasured outcomes or proposed future measures as established pilot results.</li>\n</ul>\n\n<h2>Completion Verification</h2>\n<p>Before making a recommendation, verify that the record is complete and internally consistent:</p>\n<ul>\n<li>☐ Confirm the summary retains the 24-volunteer cohort, four-week period, checklist exposure for all participants, and absence of a comparison group.</li>\n<li>☐ Recheck the support-request counts, units, subtraction, and percentage calculation.</li>\n<li>☐ Confirm that all six PILOT-03 evidence gaps and the uncontrolled exposure and task mix are visible.</li>\n<li>☐ Check that every result is labeled as observed and that no unsupported causal explanation appears.</li>\n<li>☐ Verify that a proposed extension names its unresolved question, measures, and review period before additional information is collected.</li>\n<li>☐ Confirm that the proposal remains distinct from a company-wide rollout.</li>\n</ul>\n<p>If any check remains open, revise the evidence summary or clarify the proposed review before making the extension decision.</p>\n\n<h2>Next Steps After Completion</h2>\n<p>Choose the next step according to the question and evidence still missing:</p>\n<ul>\n<li>☐ <strong>Do not extend</strong> if the record is insufficient and further observation has no specific question to answer. Document the uncertainty. This does not establish that the checklist is ineffective.</li>\n<li>☐ <strong>Consider a bounded extension</strong> only after naming the gap it will address, the measures to collect, and the review period. Possible gaps include time-to-productivity, completion quality, costs, non-volunteer outcomes, seasonality, or results after week four.</li>\n<li>☐ <strong>Clarify the comparison needed</strong> if decision-makers need to assess whether the checklist contributed to a change. Define what comparison would help; the current pilot has none.</li>\n</ul>\n<p>Do not promise a particular outcome from an extension. Keep observations, interpretations, and remaining uncertainties distinct in the decision record.</p>\n\n<h2>Conclusion</h2>\n<p></p><p>The pilot recorded fewer support requests in week four than in week one. But it does not show why the numbers changed.</p>\n<p>Before recommending a next step, use the checklist to check the pilot's scope, calculations, limits, and outcomes that were not measured.</p>\n<p>Follow the checklist in order. Check off each item only after you complete it.</p><div class=\"admonition-block tip\" style=\"display: flex; gap: 16px; padding: 16px; margin: 24px 0; background: rgba(16, 185, 129, 0.1); border-left: 4px solid rgba(16, 185, 129, 1); border-radius: 8px\"><span style=\"font-size: 24px; flex-shrink: 0\">💡</span><div class=\"admonition-content\" style=\"flex: 1\"><strong style=\"color: rgba(16, 185, 129, 1); display: block; margin-bottom: 4px\">Pro Tip</strong><span style=\"color: rgba(226, 232, 240, 1)\">Before comparing the 30 and 18 support requests, verify that the reporting weeks use the same request definition. Exposure and task mix were uncontrolled, so the 40% descriptive decrease does not establish a causal effect or authorize a company-wide rollout.</span></div></div>"
}
