Jenkins Research Institute · jSMS · Rollout plan v1, October 27, 2026
Annual safety climate survey rollout
Use one annual invitation cycle and a practical follow-up process. A survey is useful when workers can understand the findings and see what happens next.
| When | Action | Evidence to keep in jSMS |
|---|---|---|
| 4–6 weeks before | Name an executive sponsor and practical coordinator. Involve worker representatives. Define the work unit, purpose, languages and planned use. Commit resources to respond to findings, including unfavorable results. | Purpose, responsibilities, worker input and resources. Z10 5.1–5.2 and 6. |
| 3–4 weeks before | Read the current consent and no-oversight disclosure. Assess local obligations and workforce power imbalances. Arrange private time and access across shifts. Do not pressure people or require proof of completion. | Methods/privacy review, issues and decisions. Keep consent and individual answers out of company working records. |
| 2 weeks before | Choose the company connection, annual date and U.S. time zone. Preview the current connected people across roles. Remove stale company connections through normal administration. Never merge similarly named directories or import an address book by assumption. | Selected directory, settings and review date. Pending, inactive and departed connections are excluded. |
| Annual date | Send one optional invitation to each eligible current connected person. One company cycle per local calendar year prevents repeated annual sends. A February 29 anniversary uses February 28 in non-leap years. Exact delivery hour is not guaranteed. Opt-outs and bounce/complaint suppression override sending. | Delivery administration records only. Sending an invitation is not enrollment, consent or completion. |
| Response window | Use the configured 7–30 day window, normally 14 days. Share the purpose, privacy limits, contacts and how findings will be used. People may decline, skip items or stop. Do not ask supervisors to collect responses or maintain completion lists. | Communication plan and access arrangements. There are no automatic reminder campaigns in this annual workflow. |
| After the window | Review eligible grouped findings and limitations. Preserve the seven dimension means and missingness rules. No individual results or participant-level employer data. Withhold small or identifying cells; do not promise a company-level report when the study cannot safely support one. | Authorized aggregate source/version, scope, coverage, limitations and review. Z10 9.1. |
| Within 4–6 weeks | Return understandable findings to workers. Discuss the work conditions behind a pattern. Select a few priorities; assign an owner, resources and dates. Explain what will change, what cannot yet change and why. | Decisions and improvement actions with verification plans. Z10 6, 7.3 and 9.4–9.5. |
| At 30–90 days | Check whether the agreed process or safeguard actually works. Report progress back. A completed action or a more favorable perception score alone does not prove effective protection. | Field/effectiveness evidence linked to actions and management review. Z10 9 and 10. |
| Before the next anniversary | Review scope, workforce mix, annual settings and prior decisions. Keep methods comparable where appropriate; document changes. Allow interventions time to settle before interpreting a repeat survey. Annual invitations and quarterly industry publication serve different purposes. | Management-review decisions and the next plan. Z10 10.1–10.2. |
Suggested announcement
“We are inviting everyone currently connected to our company to consider a voluntary safety climate survey. We want to learn how safety is experienced and identify conditions to improve. Taking part is your choice. Please read JRI’s participant information, including the absence of independent ethics review and privacy limits. Individual answers and completion lists are not provided to us. We will explain the eligible grouped findings and the decisions we make in response.”
Keep invitation administration separate
Company managers can review their connected recipient directory and delivery administration. The survey does not receive company, account or invitation identifiers. The public collection cannot verify the number of distinct workplaces or provide linked company-specific results. Delivery records do not tell a manager who answered. An open public survey also cannot guarantee that every individual submits only once; this limitation must remain visible in reporting.
Sources and limits
NFA use guidance emphasizes voluntary participation, action and returning findings; it suggests allowing one to two years when repeating a survey to assess interventions. This rollout schedule is JRI’s practical guidance, not an NFA requirement or a Z10 conformity assessment. Read the full participant information.
Quarterly collections and optional company totals
Separate survey collections run by calendar quarter in Eastern time. Each closes as the next begins; annual invitations remain once per company per year. A requested response window never extends a closed quarter. Employees may answer once per workplace per quarter with fresh consent. This cadence does not establish that enough time has passed to evaluate an intervention.
After submitting, an actively connected employee can separately choose the company participation counter. Do not require it or ask employees to prove their participation. Employers receive no live counts, names, individual status, answer results or demographic breakdowns. After close and a seven-day buffer, only protected totals may be shown. Counts below 20 or leaving fewer than 5 eligible nonparticipants are withheld. Totals include only voluntary counted contributions, not everyone who answered publicly. Provide workers the same understandable participation update and explain resulting actions.