Day-30 Pulse — Instrument Spec
The survey that feeds row 2 of the Q1 board report. Armed in W4, fires 30 days after each participant's first session. Per-participant, not per-cohort, so the Q1 report has sample size.
Why per-participant timing
If everyone in the cohort gets surveyed on the same calendar day, half are 10 days post-first-session and half are 40 days post. The signal blurs. Per-participant timing keeps every response at the same dose.
Survey design rules
- 8 questions. No more. Anything past Q8 has a measurable response-rate cliff.
- Mobile-first. 90% of responses come on phones.
- Anonymous by default. The participant can choose to be named for follow-up, but the default is unnamed.
- One open-text box at the end, not throughout. Sprinkled text boxes kill completion rates.
- No leading wording. "How much did you love the session?" is banned. "How would you rate the session?" is fine.
The 8 questions
Q1 — Programme experience (1–5)
Overall, how would you rate the wellbeing programme so far?
Scale: 1 (poor) – 5 (excellent)
Q2 — Manager confidence (only on manager-cohort programmes; 1–5)
Thinking about your team, how confident do you feel having a supportive conversation if you spotted an early sign of pressure?
Scale: 1 (not confident) – 5 (very confident)
Paired with the same Q asked at D1 (pre) to give the pre→post delta that lands in row 2 of the Q1 report.
Q3 — Behaviour: conversations opened (yes/no/numeric)
In the last 30 days, have you opened a supportive conversation with a direct report based on what you noticed?
Options: No · Yes — once · Yes — 2 or 3 times · Yes — more than 3 times
Q4 — Behaviour: adjustments made (yes/no/numeric)
In the last 30 days, have you made a workload, recovery or scheduling adjustment for a direct report based on what you noticed?
Options: No · Yes — one · Yes — more than one
Q5 — Signal recognition (multi-select)
Which of the four signals did you spot in any direct report in the last 30 days? (Tick all that apply.)
- Energy drift
- Scope shrink
- Recovery gap
- Relational fade
- None of these
Q6 — Pocket-card use (yes/no)
Is your pocket-card still where you can see it?
Options: Yes — on my monitor or desk · Yes — but in a drawer · No · Lost it
(The "lost it" answer triggers an auto-trigger to send a replacement PDF.)
Q7 — Signposting (yes/no)
In the last 30 days, did you signpost any direct report to EAP, GP, the MEM coach, or 116 123?
Options: No · Yes — once · Yes — more than once
Q8 — Open text (optional)
Anything you want us to know? (One sentence is fine.)
Open text, optional, 500-char limit.
Variants by cohort
| Cohort | Use these Qs |
|---|---|
| Manager mental-health programme | Q1, Q2, Q3, Q4, Q5, Q6, Q7, Q8 (above) |
| Menopause manager training | Q1, Q2 (re-worded for menopause), Q3, Q4 (adjustment-focused), Q7 (signposting), Q8 |
| All-staff financial wellbeing | Q1, Q2 (re-worded: confidence with money basics), open Q on which strand helped most, Q7, Q8 |
| Hybrid wellbeing | Q1, Q2 (re-worded: confidence with rituals), Q3, Q4 (ritual adopted), Q8 |
Q1 + Q8 are constant across every variant. Q2 always tests confidence pre→post.
Operations
- Stored in: platform survey scheduler.
- Trigger: 30 days after the participant's first session attendance is logged.
- Channel: email primary, SMS fallback after 5 days no-response.
- Reminder: ONE reminder at day 7. No more.
- Close: 14 days after first send.
- Anonymity: responses stored against a hashed participant ID; the human-readable name is never joined to the response unless the participant opts to be named in the final field.
- Right to delete: any participant may delete their response within 30 days of submission; standard data-rights process.
Analysis rubric
For each cohort:
| Metric | How it's calculated |
|---|---|
| Manager-level satisfaction | Mean of Q1 |
| Confidence pre→post | Mean(Q2 post) − Mean(Q2 pre, from D1 baseline) |
| % who opened ≥1 conversation | Q3 ≠ "No" |
| % who made ≥1 adjustment | Q4 ≠ "No" |
| Signal-spotting rate | Q5 — distribution across the four signals |
| Pocket-card adherence | % "Yes — on my monitor or desk" in Q6 |
| Signposting rate | Q7 ≠ "No" |
| Open-text themes | Coded by MEM PL into 5 themes max per quarter |
Quality bar (MEM PL self-check before Q1 report)
- Response rate ≥ 50% of participants. If lower, flag in the appendix and propose remediation.
- No theme in open text appears only once with high emotional weight without being read by the Programme Lead in full.
- Confidence delta is rounded to 1 decimal place. No false precision.
- Numbers cross-checked against the delivery-system attendance data.
Reporting
The analysed pulse data feeds:
- Row 2 of the Q1 board report (the four indicators)
- The "named risk + named win" paragraph (sourced from open text)
- The recommendations on row 5 (sourced from signal-distribution + signposting-rate)
What's deliberately NOT in this instrument
- A "would you recommend MEM to a friend?" NPS-style question — vanity metric, not actionable
- A clinical screen (PHQ-9, GAD-7, etc.) — we are not a clinical provider; using a validated clinical screen without clinical follow-up is harmful
- A satisfaction-with-coach rating — handled separately in the coach feedback loop, not in the cross-cohort instrument
- A free-text comment box at every question — kills completion rate
Cross-cutting promise Every employee seat funds a free seat for someone leaving prison — the SROI ratio computed from this instrument lands in row 3 of the Q1 board report. memacademy.org/corporate/programme