Home › Intelligence › Population Health

Population Health Dashboard

Your HRIS Tells You Where They Are.
Bearings Tells You How They're Doing.

Live cultural adjustment status for every active assignee — by domain, by week, by destination. Updated every 6 hours. Based entirely on platform engagement behaviour, never on self-reported surveys.

Book a Demo → See Early Warning Engine

The Snapshot Architecture

Every 6 Hours. Every Assignee. Every Domain.

The AssigneeSnapshot entity stores a point-in-time record of every active assignee's status: 6 cultural domain scores, engagement rate, engagement trend, partner enrollment, sentiment trend, milestone completion, and risk level. Every 6 hours, a new row is added — creating a time series that enables trend analysis no other tool provides.

6hrs
Snapshot refresh cadence — new row inserted for every active assignee
6
Cultural domains tracked per assignee: Workplace, Social, Language, Family, Healthcare, Safety
12+
Weeks of engagement trend data before the Early Warning Engine achieves high-confidence predictions
47
Fields stored per snapshot — enabling cross-stream pattern detection no single field reveals

Engagement as a Leading Indicator

The Signal That Precedes Everything Else

Declining platform engagement typically precedes declining sentiment by 2–3 weeks. Declining sentiment precedes an early return request by 6–8 weeks. Engagement is the earliest signal in the system — and it is entirely behavioural, not self-reported.

An assignee who was highly active for 8 weeks and then went quiet is a different pattern from one who never engaged consistently. The snapshot stores both — and the Early Warning Engine knows the difference.

  • engagement_rate: proportion of last 4 weeks with at least one meaningful platform interaction
  • engagement_trend: computed from weeks 1–4 vs. weeks 5–8 — rising / stable / declining
  • missions_completed vs. missions_expected: are they keeping pace with the coaching cadence?
  • last_active: when did they last open the platform — a week gap is different from a day gap
Engagement Analysis — Marcus R. · Tokyo · Week 9
28%
Engagement Rate
Was: 90%
▼
Trend
Declining 3 wks
8-Week Engagement History
▼ Sharp drop at week 7 — consistent with post-honeymoon pattern
Domain Progress (Current)
Workplace
72%
Social
29%
Language
44%
Family
38%
Healthcare
68%

Partner Visibility

The Most Overlooked Risk Signal in Global Mobility

73% of accompanying partners receive no cultural support from their employer. Partner disengagement is one of the strongest predictors of assignment failure. The Population Health Dashboard shows partner enrollment status, engagement rate, and sentiment trend alongside the employee — because the household is the unit of risk, not the individual.

  • partner_enrolled flag surfaced on every assignee card in the population grid
  • Assignees with unenrolled partners after week 6 surface in the Early Warning Engine (EW-002)
  • Partner engagement rate and sentiment trend shown in the Assignee Detail partner tab
  • Partner data is private — aggregate metrics only, never individual coaching content
  • 2.4× higher completion rate for assignees whose partners are enrolled vs. not enrolled
See How Partner Risk Flags Work →
Partner Status — Population View
Sarah K. — Amsterdam Wk 8 · Partner not enrolled · Family 31%
Chen L. — Dubai Wk 12 · Partner inactive 3wks · Family 28%
David M. — Zurich Wk 14 · Partner engaged · Family 71%
Priya S. — Singapore Wk 6 · Partner enrolled Week 2 · Family 64%
Insight: Partner enrollment rate for this cohort: 58%. Platform benchmark: 74%. Recommend partner enrollment campaign for the 3 unenrolled families.

Destination Portfolio View

"How Are All My Japan Assignees Doing?" — Now Answerable in Seconds

Switch from individual assignees to destination cohorts. Aggregate cultural progress, engagement rate, risk distribution, and partner enrollment rate for every destination in your programme — compared against the platform benchmark for that destination.

🗾
Japan Cohort (8 assignees)
Avg progress: 64%. Benchmark: 58%. Social domain: 31% — below benchmark at 41%. 2 at-risk flags active. Partner enrollment: 62%.
🇩🇪
Germany Cohort (5 assignees)
Avg progress: 77%. Benchmark: 71%. Outperforming on Workplace (82%) and Language (68%). No active risk flags. Partner enrollment: 80%.
🇦🇪
UAE Cohort (4 assignees)
Avg progress: 59%. Benchmark: 63%. Family domain underperforming (41% vs benchmark 54%). 1 medium risk flag. Partner enrollment: 50%.

See Your Full Population in One View.

We'll show you a live population dashboard built for your destination mix — with real engagement data, domain progress charts, and milestone tracking.

Book a Team Demo → See Pricing

FAQ

Common Questions

How often is the dashboard updated?
The AssigneeSnapshot is refreshed every 6 hours. All domain scores, engagement rates, and risk levels reflect data within the last 6 hours — not yesterday's HRIS export.
Can I see individual coaching conversation content?
No. The dashboard shows behavioural metrics: engagement rate, mission completion, domain progress, check-in sentiment trends. Individual coaching conversations are private to the employee at the query layer — not just hidden in the UI.
What is a cultural domain score?
A composite score (0–100%) computed from mission performance in that domain, mentor interaction depth on domain-specific topics, and check-in sentiment for domain-related questions. Not self-reported. Updated with every platform interaction.
What is engagement rate?
The proportion of the last 4 calendar weeks in which the assignee had at least one meaningful platform interaction: mission attempt, mentor session, check-in response, or calendar coaching interaction.
Can I filter the population?
Yes — by destination country, assignment phase, risk level, assignment type, and partner enrollment status. Filters can be saved as named views and shared with colleagues.
Does the benchmark show other clients' data?
Never. Benchmarks are computed from anonymised aggregate data — no individual or organisation is identifiable. Your cohort data appears as a distinct line vs. the anonymised benchmark.