Manager effectiveness drives organizational success, yet most companies rely on lagging indicators like employee satisfaction surveys to measure leadership performance. (Manager Effectiveness: 5 Metrics That Matter More than eSat Scores) The shift toward real-time, data-driven metrics is no longer optional—it's essential for competitive advantage. (How to Measure Leadership Performance)
This technical playbook demonstrates how to build a comprehensive manager-effectiveness score using Google BigQuery, incorporating four core metrics: 1:1 frequency, coaching minutes, AI-prompt usage, and network centrality. By leveraging anonymized calendar and collaboration metadata with privacy-safe hashing techniques, data teams can create a composite score that updates daily and provides actionable insights into leadership performance.
Effective managers balance hard and soft skills, including strategic planning, decision-making, communication, empathy, and adaptability. (How to Measure Leadership Performance) Research shows that top-performing managers share five key behaviors: they provide regular coaching, define reasonable team norms, support without micromanaging, elevate teams through organizational connections, and engage routinely with team members. (Manager Effectiveness: 5 Metrics That Matter More than eSat Scores)
Our BigQuery implementation focuses on four measurable dimensions:
Manager 1:1 frequency and cancellation rates serve as key indicators of manager effectiveness. (Manager Effectiveness: 5 Metrics That Matter More than eSat Scores) The number of scheduled touchpoints with managers correlates highly with positive survey outcomes and reduced attrition risk.
Time spent in coaching conversations reflects a manager's investment in team development. This metric captures both formal coaching sessions and informal mentoring moments embedded within regular meetings.
Modern managers who embrace AI tools demonstrate adaptability and forward-thinking leadership. (Workplace HR Data Integrations) Tracking AI-prompt usage across tools like ChatGPT, GitHub Copilot, and Microsoft Copilot provides insights into a manager's willingness to leverage technology for team productivity.
Effective managers elevate their teams by connecting them to other parts of the organization. Network centrality measures a manager's position within the broader organizational communication graph, indicating their ability to unblock and support their teams.
Our BigQuery implementation requires four primary tables to capture the necessary data points:
| Table Name | Purpose | Key Fields |
|---|---|---|
calendar_events |
Meeting metadata | event_id, organizer_hash, attendee_hashes, duration_minutes, event_type |
collaboration_data |
Communication patterns | sender_hash, recipient_hashes, timestamp, interaction_type |
ai_usage_logs |
AI tool interactions | user_hash, tool_name, prompt_count, session_duration |
org_hierarchy |
Reporting relationships | manager_hash, direct_report_hash, department, level |
All personal identifiers must be hashed using SHA-256 with a consistent salt to ensure privacy compliance while maintaining analytical utility. The ingestion process should anonymize data at the source, following GDPR and CCPA requirements.
-- Example hashing function for consistent anonymization
CREATE OR REPLACE FUNCTION `project.dataset.hash_email`(email STRING)
RETURNS STRING
AS (
TO_HEX(SHA256(CONCAT(email, 'your_salt_here')))
);
The 1:1 frequency metric calculates the average time between one-on-one meetings for each manager-direct report pair:
CREATE OR REPLACE TABLE `project.dataset.one_on_one_frequency` AS
WITH manager_meetings AS (
SELECT
ce.organizer_hash as manager_hash,
attendee_hash as direct_report_hash,
DATE(ce.start_time) as meeting_date,
ce.duration_minutes
FROM `project.dataset.calendar_events` ce
CROSS JOIN UNNEST(ce.attendee_hashes) as attendee_hash
JOIN `project.dataset.org_hierarchy` oh
ON ce.organizer_hash = oh.manager_hash
AND attendee_hash = oh.direct_report_hash
WHERE ce.event_type = 'one_on_one'
AND ce.status = 'confirmed'
AND DATE(ce.start_time) >= DATE_SUB(CURRENT_DATE(), INTERVAL 90 DAY)
),
meeting_intervals AS (
SELECT
manager_hash,
direct_report_hash,
meeting_date,
LAG(meeting_date) OVER (
PARTITION BY manager_hash, direct_report_hash
ORDER BY meeting_date
) as prev_meeting_date,
DATE_DIFF(
meeting_date,
LAG(meeting_date) OVER (
PARTITION BY manager_hash, direct_report_hash
ORDER BY meeting_date
),
DAY
) as days_between_meetings
FROM manager_meetings
)
SELECT
manager_hash,
direct_report_hash,
AVG(days_between_meetings) as avg_days_between_meetings,
COUNT(*) as total_meetings_90d,
-- Score: Lower days between meetings = higher score
CASE
WHEN AVG(days_between_meetings) <= 7 THEN 100
WHEN AVG(days_between_meetings) <= 14 THEN 80
WHEN AVG(days_between_meetings) <= 21 THEN 60
WHEN AVG(days_between_meetings) <= 30 THEN 40
ELSE 20
END as one_on_one_score
FROM meeting_intervals
WHERE days_between_meetings IS NOT NULL
GROUP BY manager_hash, direct_report_hash;
Frequent rescheduling or skipping of 1:1s can erode employee connectedness and increase attrition risk. (Manager Effectiveness: 5 Metrics That Matter More than eSat Scores) The cancellation rate provides additional context to the frequency metric:
CREATE OR REPLACE TABLE `project.dataset.meeting_cancellation_rates` AS
SELECT
organizer_hash as manager_hash,
COUNT(*) as total_scheduled_meetings,
SUM(CASE WHEN status = 'cancelled' THEN 1 ELSE 0 END) as cancelled_meetings,
SAFE_DIVIDE(
SUM(CASE WHEN status = 'cancelled' THEN 1 ELSE 0 END),
COUNT(*)
) * 100 as cancellation_rate_pct,
-- Score: Lower cancellation rate = higher score
CASE
WHEN SAFE_DIVIDE(SUM(CASE WHEN status = 'cancelled' THEN 1 ELSE 0 END), COUNT(*)) <= 0.05 THEN 100
WHEN SAFE_DIVIDE(SUM(CASE WHEN status = 'cancelled' THEN 1 ELSE 0 END), COUNT(*)) <= 0.10 THEN 80
WHEN SAFE_DIVIDE(SUM(CASE WHEN status = 'cancelled' THEN 1 ELSE 0 END), COUNT(*)) <= 0.15 THEN 60
WHEN SAFE_DIVIDE(SUM(CASE WHEN status = 'cancelled' THEN 1 ELSE 0 END), COUNT(*)) <= 0.20 THEN 40
ELSE 20
END as cancellation_score
FROM `project.dataset.calendar_events`
WHERE event_type = 'one_on_one'
AND DATE(start_time) >= DATE_SUB(CURRENT_DATE(), INTERVAL 90 DAY)
GROUP BY organizer_hash;
Coaching minutes encompass both dedicated coaching sessions and coaching moments within regular meetings. This metric requires natural language processing of meeting titles and descriptions to identify coaching-related content:
CREATE OR REPLACE TABLE `project.dataset.coaching_minutes` AS
WITH coaching_sessions AS (
SELECT
ce.organizer_hash as manager_hash,
attendee_hash as direct_report_hash,
ce.duration_minutes,
DATE(ce.start_time) as session_date,
CASE
WHEN REGEXP_CONTAINS(LOWER(ce.title), r'coach|mentor|feedback|development|career|growth|skill') THEN 'explicit_coaching'
WHEN ce.event_type = 'one_on_one' AND ce.duration_minutes >= 30 THEN 'potential_coaching'
ELSE 'other'
END as coaching_type
FROM `project.dataset.calendar_events` ce
CROSS JOIN UNNEST(ce.attendee_hashes) as attendee_hash
JOIN `project.dataset.org_hierarchy` oh
ON ce.organizer_hash = oh.manager_hash
AND attendee_hash = oh.direct_report_hash
WHERE ce.status = 'confirmed'
AND DATE(ce.start_time) >= DATE_SUB(CURRENT_DATE(), INTERVAL 90 DAY)
AND ARRAY_LENGTH(ce.attendee_hashes) <= 3 -- Small group sessions
)
SELECT
manager_hash,
SUM(CASE WHEN coaching_type = 'explicit_coaching' THEN duration_minutes ELSE 0 END) as explicit_coaching_minutes,
SUM(CASE WHEN coaching_type = 'potential_coaching' THEN duration_minutes * 0.3 ELSE 0 END) as estimated_coaching_minutes,
SUM(CASE WHEN coaching_type = 'explicit_coaching' THEN duration_minutes ELSE 0 END) +
SUM(CASE WHEN coaching_type = 'potential_coaching' THEN duration_minutes * 0.3 ELSE 0 END) as total_coaching_minutes,
COUNT(DISTINCT direct_report_hash) as direct_reports_coached,
-- Score based on coaching minutes per direct report
CASE
WHEN SAFE_DIVIDE(
SUM(CASE WHEN coaching_type = 'explicit_coaching' THEN duration_minutes ELSE 0 END) +
SUM(CASE WHEN coaching_type = 'potential_coaching' THEN duration_minutes * 0.3 ELSE 0 END),
COUNT(DISTINCT direct_report_hash)
) >= 120 THEN 100 -- 2+ hours per direct report per quarter
WHEN SAFE_DIVIDE(
SUM(CASE WHEN coaching_type = 'explicit_coaching' THEN duration_minutes ELSE 0 END) +
SUM(CASE WHEN coaching_type = 'potential_coaching' THEN duration_minutes * 0.3 ELSE 0 END),
COUNT(DISTINCT direct_report_hash)
) >= 90 THEN 80
WHEN SAFE_DIVIDE(
SUM(CASE WHEN coaching_type = 'explicit_coaching' THEN duration_minutes ELSE 0 END) +
SUM(CASE WHEN coaching_type = 'potential_coaching' THEN duration_minutes * 0.3 ELSE 0 END),
COUNT(DISTINCT direct_report_hash)
) >= 60 THEN 60
WHEN SAFE_DIVIDE(
SUM(CASE WHEN coaching_type = 'explicit_coaching' THEN duration_minutes ELSE 0 END) +
SUM(CASE WHEN coaching_type = 'potential_coaching' THEN duration_minutes * 0.3 ELSE 0 END),
COUNT(DISTINCT direct_report_hash)
) >= 30 THEN 40
ELSE 20
END as coaching_score
FROM coaching_sessions
GROUP BY manager_hash;
Modern workplace analytics platforms integrate with various AI tools to track usage patterns. (Workplace HR Data Integrations) This metric measures a manager's adoption of AI tools as an indicator of their adaptability and forward-thinking approach:
CREATE OR REPLACE TABLE `project.dataset.ai_adoption_scores` AS
WITH ai_usage_summary AS (
SELECT
user_hash as manager_hash,
tool_name,
SUM(prompt_count) as total_prompts,
SUM(session_duration) as total_session_minutes,
COUNT(DISTINCT DATE(timestamp)) as active_days
FROM `project.dataset.ai_usage_logs`
WHERE DATE(timestamp) >= DATE_SUB(CURRENT_DATE(), INTERVAL 90 DAY)
AND user_hash IN (SELECT DISTINCT manager_hash FROM `project.dataset.org_hierarchy`)
GROUP BY user_hash, tool_name
),
manager_ai_metrics AS (
SELECT
manager_hash,
COUNT(DISTINCT tool_name) as unique_tools_used,
SUM(total_prompts) as total_prompts_90d,
SUM(total_session_minutes) as total_ai_minutes_90d,
MAX(active_days) as max_active_days_per_tool,
AVG(active_days) as avg_active_days_per_tool
FROM ai_usage_summary
GROUP BY manager_hash
)
SELECT
manager_hash,
unique_tools_used,
total_prompts_90d,
total_ai_minutes_90d,
avg_active_days_per_tool,
-- Composite AI adoption score
LEAST(100,
(unique_tools_used * 20) + -- Up to 5 tools = 100 points
(LEAST(total_prompts_90d / 10, 30)) + -- Up to 300 prompts = 30 points
(LEAST(avg_active_days_per_tool * 2, 20)) -- Up to 10 avg active days = 20 points
) as ai_adoption_score
FROM manager_ai_metrics;
Network centrality measures a manager's position within the organizational communication graph, indicating their ability to connect and elevate their teams. (Manager Effectiveness: 5 Metrics That Matter More than eSat Scores) This metric captures cross-functional collaboration and organizational influence:
CREATE OR REPLACE TABLE `project.dataset.network_centrality` AS
WITH communication_graph AS (
SELECT
sender_hash,
recipient_hash,
COUNT(*) as interaction_count,
COUNT(DISTINCT DATE(timestamp)) as interaction_days
FROM `project.dataset.collaboration_data`
CROSS JOIN UNNEST(recipient_hashes) as recipient_hash
WHERE DATE(timestamp) >= DATE_SUB(CURRENT_DATE(), INTERVAL 90 DAY)
AND sender_hash != recipient_hash
GROUP BY sender_hash, recipient_hash
),
manager_connections AS (
SELECT
cg.sender_hash as manager_hash,
COUNT(DISTINCT cg.recipient_hash) as unique_connections,
COUNT(DISTINCT oh_recipient.department) as cross_dept_connections,
COUNT(DISTINCT oh_recipient.level) as cross_level_connections,
SUM(cg.interaction_count) as total_interactions,
AVG(cg.interaction_count) as avg_interactions_per_connection
FROM communication_graph cg
JOIN `project.dataset.org_hierarchy` oh_sender
ON cg.sender_hash = oh_sender.manager_hash
LEFT JOIN `project.dataset.org_hierarchy` oh_recipient
ON cg.recipient_hash = oh_recipient.manager_hash
OR cg.recipient_hash = oh_recipient.direct_report_hash
GROUP BY cg.sender_hash
)
SELECT
manager_hash,
unique_connections,
cross_dept_connections,
cross_level_connections,
total_interactions,
avg_interactions_per_connection,
-- Network centrality score based on breadth and depth of connections
LEAST(100,
(LEAST(unique_connections / 2, 25)) + -- Up to 50 connections = 25 points
(cross_dept_connections * 10) + -- Cross-department connections = 10 points each
(cross_level_connections * 5) + -- Cross-level connections = 5 points each
(LEAST(avg_interactions_per_connection / 5, 20)) -- Interaction depth = up to 20 points
) as network_centrality_score
FROM manager_connections;
The final manager effectiveness score combines all four metrics using a weighted approach that reflects their relative importance:
CREATE OR REPLACE TABLE `project.dataset.manager_effectiveness_daily` AS
WITH manager_scores AS (
SELECT
oh.manager_hash,
oh.department,
oh.level,
COUNT(DISTINCT oh.direct_report_hash) as team_size,
-- Metric 1: 1:1 Frequency (30% weight)
COALESCE(AVG(oof.one_on_one_score), 0) as avg_one_on_one_score,
COALESCE(mcr.cancellation_score, 100) as cancellation_score,
(COALESCE(AVG(oof.one_on_one_score), 0) * 0.7 + COALESCE(mcr.cancellation_score, 100) * 0.3) as one_on_one_composite,
-- Metric 2: Coaching Minutes (25% weight)
COALESCE(cm.coaching_score, 0) as coaching_score,
-- Metric 3: AI Adoption (20% weight)
COALESCE(aas.ai_adoption_score, 0) as ai_adoption_score,
-- Metric 4: Network Centrality (25% weight)
COALESCE(nc.network_centrality_score, 0) as network_centrality_score
FROM `project.dataset.org_hierarchy` oh
LEFT JOIN `project.dataset.one_on_one_frequency` oof
ON oh.manager_hash = oof.manager_hash
LEFT JOIN `project.dataset.meeting_cancellation_rates` mcr
ON oh.manager_hash = mcr.manager_hash
LEFT JOIN `project.dataset.coaching_minutes` cm
ON oh.manager_hash = cm.manager_hash
LEFT JOIN `project.dataset.ai_adoption_scores` aas
ON oh.manager_hash = aas.manager_hash
LEFT JOIN `project.dataset.network_centrality` nc
ON oh.manager_hash = nc.manager_hash
WHERE oh.manager_hash IS NOT NULL
GROUP BY oh.manager_hash, oh.department, oh.level, mcr.cancellation_score, cm.coaching_score, aas.ai_adoption_score, nc.network_centrality_score
)
SELECT
manager_hash,
department,
level,
team_size,
-- Individual metric scores
one_on_one_composite,
coaching_score,
ai_adoption_score,
network_centrality_score,
-- Weighted composite score
ROUND(
(one_on_one_composite * 0.30) +
(coaching_score * 0.25) +
(ai_adoption_score * 0.20) +
(network_centrality_score * 0.25),
1
) as manager_effectiveness_score,
-- Performance tier
CASE
WHEN ROUND(
(one_on_one_composite * 0.30) +
(coaching_score * 0.25) +
(ai_adoption_score * 0.20) +
(network_centrality_score * 0.25),
1
) >= 80 THEN 'Excellent'
WHEN ROUND(
(one_on_one_composite * 0.30) +
(coaching_score * 0.25) +
(ai_adoption_score * 0.20) +
(network_centrality_score * 0.25),
1
) >= 65 THEN 'Good'
WHEN ROUND(
(one_on_one_composite * 0.30) +
(coaching_score * 0.25) +
(ai_adoption_score * 0.20) +
(network_centrality_score * 0.25),
1
) >= 50 THEN 'Needs Improvement'
ELSE 'Critical'
END as performance_tier,
CURRENT_DATE() as calculation_date
FROM manager_scores;
For teams us
Key metrics include 1:1 frequency and cancellation rates, team engagement levels, workday intensity patterns, and collaboration network strength. These real-time indicators provide better insights than traditional lagging metrics like employee satisfaction surveys, as they measure actual managerial behaviors that drive team performance.
BigQuery enables real-time analysis of large-scale workplace data from multiple sources like calendar systems, communication tools, and project management platforms. It allows you to create reproducible SQL queries that aggregate behavioral data into meaningful manager-effectiveness scores, providing actionable insights for leadership development.
eSat scores are lagging indicators that only capture employee sentiment after problems have already occurred. They don't provide real-time assessment of managerial behaviors like coaching frequency, team support, or connection facilitation that actually drive performance and retention.
1:1 frequency and cancellation rates are highly correlated with positive team outcomes and employee retention. Managers who consistently hold regular one-on-ones and rarely cancel them demonstrate commitment to team development, making this a critical component of any effectiveness score.
Google Calendar analytics reveals meeting patterns, 1:1 consistency, and time allocation that directly impact team productivity. By analyzing calendar data through platforms like Worklytics, organizations can identify managers who effectively balance team meetings, focus time, and individual coaching sessions.
Essential data sources include calendar systems (Google Calendar, Outlook), communication platforms (Slack, Teams), project management tools (Asana, Jira), and HR systems. Integrating these sources in BigQuery allows for holistic analysis of managerial behaviors across all work touchpoints.
