Worklytics provides datasets describing work at various levels of aggregation. From less to more granular, the key Worklytics datasets are as follows:

*Note: Employee Events & Work Items datasets are only available in Datastream Enterprise tiers. Please contact sales@worklytics.co for more information

The following is a sample set of SQL queries you can use to get started with the Worklytics dataset. This code block assumes the user has access to the individual aggregates, individual groups, and collaboration graph datasets, though the exact table names will have to be substituted for the ones in your company's environment.

/* 
SAMPLE SQL QUERIES FOR AGGREGATE-LEVEL DATA
DATE: 2024.09.19

INPUT TABLES: 
{weekly_individual_aggregates}
{weekly_individual_groups}
{collaboration_graph}
{collaboration_graph_split}


SECTION 1: IDENTIFY KEY METRICS ABOUT YOUR ORGANIZATION'S MEETING CULTURE, FOCUS TIME, WORK-LIFE BALANCE AND COLLABORATION CULTURE USING THE INDIVIDUAL AGGREGATES TABLE
SECTION 2: IDENTIFY THE TYPICAL NUMBER OF COLLABORATORS (AND STRONG COLLABORATORS) AND TYPICAL NUMBER OF SLACK COLLABORATORS

NOTE: QUERIES HAVE BEEN WRITTEN FOR USE IN BIGQUERY - THEREFORE CERTAIN FUNCTIONS MAY NEED TO BE ADJUSTED FOR USE IN ANOTHER ENVIRONMENT

*/

------------------------------------------------------------------------------------------------------------------------------
--QUERY 1.1: IDENTIFY KEY METRICS ABOUT YOUR ORGANIZATION'S MEETING CULTURE
--INCLUDES METRICS FOR MEDIAN NUMBER OF WEEKLY MEETING HOURS, % OF MEETING TIME FOCUSED AND DAILY FOCUS HOURS
--NOTE THAT METRICS INCLUDE ONLY ACTIVE EMPLOYEES THAT WORKED A MINIMUM OF 4 HOURS EACH DAY AND INCLUDES ONLY FULL-TIME EMPLOYEES
------------------------------------------------------------------------------------------------------------------------------
WITH active_min_hours_emps AS (
SELECT DISTINCT a.week, a.employeeId
FROM {weekly_individual_aggregates} a
JOIN {weekly_individual_groups} g 
ON a.employeeId = g.employeeId 
AND a.week = g.week
WHERE 1=1
AND g.week >= '2024-01-01' AND g.week < '2024-03-31'
AND g.custom_group_1 IN ('Full-Time') -- exact filter will vary by company
AND g.active = true
AND (a.key = 'worklytics:weekdays:avg:timespan:hours' AND a.value >= 4) 
),
weekly_totals AS (
SELECT p.week, 
       p.employeeId, 
       CAST(SUM(IF(key='calendar:events:hours:meetings',value,0)) AS NUMERIC) AS weekly_meeting_hours,
       1.0*CAST(SUM(IF(key='calendar:events:focus',value,0)) AS NUMERIC)/100.0 AS percent_meeting_focused,      
       CAST(SUM(IF(key='worklytics:hours:in:focus:blocks:v3:flow',value,0)) AS NUMERIC) AS daily_focus_hours             
FROM {weekly_individual_aggregates} p 
JOIN active_min_hours_emps e 
ON p.employeeId = e.employeeId 
AND p.week = e. week
WHERE 1=1
AND key IN ('calendar:events:hours:meetings','calendar:events:focus','worklytics:hours:in:focus:blocks:v3:flow')
GROUP BY 1,2 
),
average_totals AS (
SELECT employeeId, 
       CAST(AVG(weekly_meeting_hours) AS NUMERIC) AS average_weekly_meeting_hours,
       CAST(AVG(percent_meeting_focused) AS NUMERIC) AS average_percent_meeting_focused,
       CAST(AVG(daily_focus_hours) AS NUMERIC) AS average_daily_focus_hours
FROM weekly_totals
GROUP BY 1
)
SELECT ROUND(PERCENTILE_CONT(average_weekly_meeting_hours,0.5) OVER (),1) AS weekly_meeting_hours,
       ROUND(PERCENTILE_CONT(average_percent_meeting_focused,0.5) OVER (),2) AS percent_meeting_focused,
       ROUND(PERCENTILE_CONT(average_daily_focus_hours,0.5) OVER (),1) AS daily_focus_hours
FROM average_totals
LIMIT 1; 


------------------------------------------------------------------------------------------------------------------------------
--QUERY 1.2: IDENTIFY KEY METRICS ABOUT YOUR ORGANIZATION'S WORK-LIFE BALANCE BY LEVEL OF EMPLOYEE
--INCLUDES METRICS FOR MEDIAN NUMBER OF DAILY WORK HOURS AND WORK HOURS PER WEEKEND 
--NOTE THAT METRICS INCLUDE ONLY ACTIVE EMPLOYEES THAT WORKED A MINIMUM OF 4 HOURS EACH DAY AND INCLUDES ONLY FULL-TIME EMPLOYEES
------------------------------------------------------------------------------------------------------------------------------
WITH active_min_hours_emps AS (
SELECT DISTINCT a.week, a.employeeId, g.level
FROM {weekly_individual_aggregates} a
JOIN {weekly_individual_groups} g 
ON a.employeeId = g.employeeId 
AND a.week = g.week
WHERE 1=1
AND g.week >= '2024-01-01' AND g.week < '2024-03-31'
AND g.custom_group_1 IN ('Full-Time') -- exact filter will vary by company
AND g.active = true
AND (a.key = 'worklytics:weekdays:avg:timespan:hours' AND a.value >= 4) 
),
weekly_totals AS (
SELECT p.week, 
       p.employeeId, 
       e.level,
       CAST(SUM(IF(key='worklytics:weekdays:avg:timespan:hours',value,0)) AS NUMERIC) AS daily_work_hours, 
       CAST(SUM(IF(key='worklytics:weekends:total:time:worked:hours',value,0)) AS NUMERIC) AS weekly_hours_weekend
FROM {weekly_individual_aggregates} p 
JOIN active_min_hours_emps e 
ON p.employeeId = e.employeeId
AND p.week = e. week
WHERE 1=1
AND key IN ('worklytics:weekdays:avg:timespan:hours','worklytics:weekends:total:time:worked:hours')
GROUP BY 1,2,3 
),
average_totals AS (
SELECT employeeId, 
       level,
       CAST(AVG(daily_work_hours) AS NUMERIC) AS average_daily_work_hours,
       CAST(AVG(weekly_hours_weekend) AS NUMERIC) AS average_weekly_hours_weekend
FROM weekly_totals
GROUP BY 1,2
)
SELECT DISTINCT level, ROUND(PERCENTILE_CONT(average_daily_work_hours,0.5) OVER (PARTITION BY level),1) AS daily_work_hours,
       ROUND(PERCENTILE_CONT(average_weekly_hours_weekend,0.5) OVER (PARTITION BY level),1) AS weekly_hours_weekend
FROM average_totals
ORDER BY 1;


------------------------------------------------------------------------------------------------------------------------------
--QUERY 1.3: IDENTIFY KEY METRICS ABOUT YOUR ORGANIZATION'S COLLABORATION CULTURE
--INCLUDES METRICS FOR MEDIAN NUMBER OF WEEKLY COLLABORATORS, WEEKLY STRONG COLLABORATORS, NUMBER OF COLLABORATORS BETWEEN LINES OF BUSINESS
--AND THE NUMBER OF HOURS SPENT COLLABORATING BETWEEN LINES OF BUSINESS
--NOTE THAT METRICS INCLUDE ONLY ACTIVE EMPLOYEES THAT WORKED A MINIMUM OF 4 HOURS EACH DAY AND INCLUDES ONLY FULL-TIME EMPLOYEES
------------------------------------------------------------------------------------------------------------------------------
WITH active_min_hours_emps AS (
SELECT DISTINCT a.week, a.employeeId
FROM {weekly_individual_aggregates} a
JOIN {weekly_individual_groups} g 
ON a.employeeId = g.employeeId 
AND a.week = g.week
WHERE 1=1
AND g.week >= '2024-01-01' AND g.week < '2024-03-31'
AND g.custom_group_1 IN ('Full-Time') -- exact filter will vary by company
AND g.active = true
AND (a.key = 'worklytics:weekdays:avg:timespan:hours' AND a.value >= 4) 
),
weekly_totals AS (
SELECT p.week, 
       p.employeeId, 
       CAST(SUM(IF(key='collaborators:count_distinct',value,0)) AS NUMERIC) AS weekly_collaborators,
       CAST(SUM(IF(key='collaborators:strong_count_distinct',value,0)) AS NUMERIC) AS weekly_strong_collaborators,
       CAST(SUM(IF(key='lineofbusiness_individual_collaboration_intrateam_hours',value,0)) AS NUMERIC) AS weekly_hours_within_lob_collaboration,
       CAST(SUM(IF(key='lineofbusiness_individual_collaboration_interteam_hours',value,0)) AS NUMERIC) AS weekly_hours_between_lob_collaboration,
       CAST(SUM(IF(key='lineofbusiness_individual_collaboration_intrateam_people_count',value,0)) AS NUMERIC) AS weekly_collaborators_within_lob,
       CAST(SUM(IF(key='lineofbusiness_individual_collaboration_interteam_people_count',value,0)) AS NUMERIC) AS weekly_collaborators_between_lob 
FROM {weekly_individual_aggregates} p 
JOIN active_min_hours_emps e 
ON p.employeeId = e.employeeId
AND p.week = e. week
WHERE 1=1
AND key IN ('collaborators:count_distinct','collaborators:strong_count_distinct','lineofbusiness_individual_collaboration_intrateam_hours','lineofbusiness_individual_collaboration_interteam_hours','lineofbusiness_individual_collaboration_intrateam_people_count','lineofbusiness_individual_collaboration_interteam_people_count')
GROUP BY 1,2 
),
average_totals AS (
SELECT employeeId, 
       CAST(AVG(weekly_collaborators) AS NUMERIC) AS average_weekly_collaborators,
       CAST(AVG(weekly_strong_collaborators) AS NUMERIC) AS average_weekly_strong_collaborators,
       CAST(AVG(weekly_hours_within_lob_collaboration) AS NUMERIC) AS average_weekly_hours_within_lob_collaboration,
       CAST(AVG(weekly_hours_between_lob_collaboration) AS NUMERIC) AS average_weekly_hours_between_lob_collaboration,
       CAST(AVG(weekly_collaborators_within_lob) AS NUMERIC) AS average_weekly_collaborators_within_lob,
       CAST(AVG(weekly_collaborators_between_lob ) AS NUMERIC) AS average_weekly_collaborators_between_lob
FROM weekly_totals
GROUP BY 1
)
SELECT ROUND(PERCENTILE_CONT(average_weekly_collaborators,0.5) OVER (),1) AS weekly_collaborators,
       ROUND(PERCENTILE_CONT(average_weekly_strong_collaborators,0.5) OVER (),1) AS weekly_strong_collaborators,
       ROUND(PERCENTILE_CONT(average_weekly_hours_within_lob_collaboration,0.5) OVER (),1) AS weekly_hours_within_lob_collaboration, 
       ROUND(PERCENTILE_CONT(average_weekly_hours_between_lob_collaboration,0.5) OVER (),1) AS weekly_hours_between_lob_collaboration, 
       ROUND(PERCENTILE_CONT(average_weekly_collaborators_within_lob,0.5) OVER (),1) AS weekly_collaborators_within_lob, 
       ROUND(PERCENTILE_CONT(average_weekly_collaborators_between_lob,0.5) OVER (),2) AS weekly_collaborators_between_lob
FROM average_totals
LIMIT 1; 


------------------------------------------------------------------------------------------------------------------------------
--QUERY 2.1 IDENTIFY THE MEDIAN NUMBER OF COLLABORATORS AND STRONG COLLABORATORS BASED ON LINE OF BUSINESS
--NOTE THAT METRICS INCLUDE ONLY ACTIVE EMPLOYEES THAT WORKED A MINIMUM OF 4 HOURS EACH DAY AND INCLUDES ONLY FULL-TIME EMPLOYEES
------------------------------------------------------------------------------------------------------------------------------
WITH active_min_hours_emps AS (
SELECT DISTINCT a.week, 
       a.employeeId,
       g.custom_group_1,
       g.line_of_business
FROM {weekly_individual_aggregates} a
JOIN {weekly_individual_groups} g 
ON a.employeeId = g.employeeId 
AND a.week = g.week
WHERE 1=1
AND g.week >= '2024-01-01' AND g.week < '2024-03-31'
AND g.custom_group_1 IN ('Full-Time') -- exact filter will vary by company
AND g.active = true
AND (a.key = 'worklytics:weekdays:avg:timespan:hours' AND a.value >= 4) 
),
collaborator_type AS (
SELECT c.week,
       e.line_of_business,
       employeeIdSource AS person_id_src,
       employeeIdTarget AS person_id_dst,
       SUM(weeklyCollaborationTimeSeconds)  AS time_spent_together_exclusive_ms,
       SUM(weeklyCollaborationTimeSeconds) >= (2 * 60 * 60) AS is_strong_connection
FROM {collaboration_graph} c
JOIN active_min_hours_emps e
ON c.employeeIdSource = e.employeeId
AND DATE_TRUNC(c.week, ISOWEEK) = e.week
WHERE 1=1
GROUP BY 1,2,3,4 
),
people_connections AS (
SELECT week,
       person_id_src AS person_id,
       line_of_business,
       COUNT(DISTINCT person_id_dst) AS connections,
       COUNT(DISTINCT(IF(is_strong_connection=true, person_id_dst,NULL))) AS strong_connections
FROM collaborator_type  
GROUP BY 1,2,3
)
SELECT DISTINCT line_of_business,
       ROUND(PERCENTILE_CONT(connections,0.5) OVER (PARTITION BY line_of_business),0) AS connections,
       ROUND(PERCENTILE_CONT(strong_connections,0.5) OVER (PARTITION BY line_of_business),0) AS strong_connections
FROM people_connections
WHERE line_of_business IS NOT NULL
ORDER BY 1;


------------------------------------------------------------------------------------------------------------------------------
--QUERY 2.2 IDENTIFY THE MEDIAN NUMBER OF SLACK COLLABORATORS FOR THE TYPICAL EMPLOYEE BASED ON LINE OF BUSINESS
--NOTE THAT METRICS INCLUDE ONLY ACTIVE EMPLOYEES THAT WORKED A MINIMUM OF 4 HOURS EACH DAY AND INCLUDES ONLY FULL-TIME EMPLOYEES
------------------------------------------------------------------------------------------------------------------------------
WITH active_min_hours_emps AS (
SELECT DISTINCT a.week, 
       a.employeeId,
       g.custom_group_1,
       g.line_of_business
FROM {weekly_individual_aggregates} a
JOIN {weekly_individual_groups} g 
ON a.employeeId = g.employeeId 
AND a.week = g.week
WHERE 1=1
AND g.week >= '2024-01-01' AND g.week < '2024-03-31'
AND g.custom_group_1 IN ('Full-Time') -- exact filter will vary by company
AND g.active = true
AND (a.key = 'worklytics:weekdays:avg:timespan:hours' AND a.value >= 4) 
),
collaborator_type AS (
SELECT c.week,
       e.line_of_business,
       employeeIdSource AS person_id_src,
       employeeIdTarget AS person_id_dst
FROM {collaboration_graph_split} c
JOIN active_min_hours_emps e
ON c.employeeIdSource = e.employeeId
AND DATE_TRUNC(c.week, ISOWEEK) = e.week
WHERE 1=1
AND sourceType = 'slack'
GROUP BY 1,2,3,4 
),
people_connections_slack AS (
SELECT week,
       person_id_src AS person_id,
       line_of_business,
       COUNT(DISTINCT person_id_dst) AS slack_connections
FROM collaborator_type  
GROUP BY 1,2,3
)
SELECT DISTINCT line_of_business,
       ROUND(PERCENTILE_CONT(slack_connections,0.5) OVER (PARTITION BY line_of_business),0) AS slack_connections
FROM people_connections_slack
WHERE line_of_business IS NOT NULL
ORDER BY 1;

Worklytics provides datasets describing work at various levels of aggregation. From less to more granular, the key Worklytics datasets you will be using are as follows:

In our Getting Started content, we use the following table names to refer to specific datasets, though the exact naming convention of tables may look different in your company's environment:

New to Worklytics? Here you will find resources intended to help you get up-to-speed as quickly as possible. The resources here assume an intermediate (or above) knowledge of relational databases and SQL specifically. Resources include:

• : an overview of available datasets and how we refer to them in this section of the documentation

• : an absolute beginner's guide to Worklytics data

• : once you've successfully run the first two queries, this section goes a bit deeper into analyses that are possible with the Individual Aggregates dataset & the collaboration graph

• : finally, delve into event-level data, which allows for more sophisticated analyses and the generation of custom metrics

Interested in answering a particular question with the Worklytics dataset? We've started and will continue to aggregate methodologies the Worklytics Customer Insights team uses frequently to perform analyses here. Guides include:

First time working with the Worklytics dataset? Try running these queries first.

1. Basic Metric Aggregation

How many strong collaborators did the typical team member have each week in August?

Sample Query Output

Note that the above query can be quickly edited to show results for any of the metrics included in the dataset and any date range.

2. Basic Collaboration Analysis

How much time did the typical team member spend collaborating in slack in August? With how many collaborators?

Sample Query Output

Note that the above query can be quickly edited to show results for any of the collaboration tools and any date range.

Notes on general methodology applied to generate .

Metric Availability

Note that not all metrics in Data Dictionary will necessarily appear in your company’s export, as some require specific data connectors (e.g., if your company has not connected GitHub, you will not see any of the github: metrics). The list of metrics may change as aggregates are added over time.

Active Employees

Worklytics uses the flag worklytics:active:employees, which has a value of 1 for weeks when a given employee is active and a value of 0 for when said employee is inactive. We define active based on whether the employee has taken any action across digital tools that is not a calendar attendance or an email auto-response (which are both considered passive actions).

Default Values & After-Hours Metrics

The default work week is Monday to Friday, and the default work day is 9:00 to 17:00. Unless you have made changes to the defaults for your company or provided work hours via an HRIS file, after-hours metrics are calculated based on these values, combined with employee calendar data to take time zone into account (e.g., an email received at 19:00 local time is considered to be after-hours; a meeting on Saturday is considered weekend work). If employees have set their calendar hours to something other than the default values, we use what the employee has set to calculate after-hours.

Exclusive Time

Worklytics algorithmically estimates the time spent on each work activity to make apples-to-apples comparisons possible across tools, a process which can be relevant for measurement of tools like Slack and email that don't automatically have an associated hourly cost like meetings do. For example, an instance of someone sending 10 rapid-fire Slack messages might be estimated to take 2 minutes of time based on default values used for Slack bounded by any activities that are happening before and after each message is sent.

Exclusive time calculations underlie metrics like focus time (used broadly) and can be seen directly in metrics following the TOOL:minutes:week format (typically used only for a few specific modeling use cases).

Meetings

Meeting Definitions

To be considered a meeting, a calendar event must meet the following criteria:

• At least 2+ human attendees (e.g., adding a meeting room to a calendar invite would not count as another attendee) that have not declined the event

• Less than 7 hours in length

We include "maybe" RSVPs as meeting attendees, because in practice, we see that employees at many companies often attend meetings they have neither declined nor accepted. In cases where meetings overlap and the RSVP is different, the higher RSVP takes precedence in the following order:

1. ACCEPTED

2. TENTATIVE

3. NEEDS ACTION

4. DECLINED

Meeting Time

For meetings specifically, Worklytics provides several layers of insight:

• Meeting count: reports on the total number of meetings

• Direct time: sum of all meeting hours, regardless of when they occur

• Exclusive meeting time: de-dupes overlapping meetings from direct time

• Exclusive time: subtracts time spent on other events that happen during meetings from meeting time

For example, consider a person who accepts the calendar invite for a team all-hands (10 people) from 10-11am on Tuesday and a 1:1 (2 people) at the same time and then sends Slack messages for 5 total minutes of time during the 10-11am hour.

• Meeting count: this person attended 2 meetings

• Direct time: this person attended 2 hours of meetings

• Exclusive meeting time: this person was in meetings for a total of 1 hour (30 minutes in each because both meetings had the same RSVP and overlap fully)

• Exclusive time: this person spent 55 minutes of time exclusively on meetings (27.5 minutes on each because 5 mins of meeting time was actually spent on Slack)

In most cases, we recommend measuring "Meeting count" or "Exclusive meeting time," though there are instances where "Exclusive time" makes more sense.

Focus Time

Worklytics has created a set of metrics intended to measure the amount of time in the day available for deep, focused work. In theory, the whole span of one’s workday could be used for this purpose, but in reality, this rarely is the case. Meetings, Slack, and email can each break up blocks that could have been used for deep work — and this distraction is measurable.

This metric has several variations:

• worklytics:hours:in:focus:blocks – daily time in blocks of 2+ hours free from meetings. Useful in determining whether meetings are disrupting deep work.

• worklytics:hours:in:focus:blocks:v2 – daily time in blocks of 2+ hours free from meetings, Slack, and email. Useful in investigating a wider set of possible disruptions to deep work.

• worklytics:hours:in:focus:blocks:v3:prep – daily time in blocks of 30+ minutes free from meetings, Slack, and email. Useful as an analysis refinement for groups like sales and customer success that consider meetings a core work activity but still likely need some time to prepare for calls during the day.

• worklytics:hours:in:focus:blocks:v3:focus – daily time in blocks of 1+ hour free from meetings, Slack, and email. Useful as an analysis refinement for groups like people managers and senior leaders that need to push forward larger projects despite a typically heavy meeting load.

• worklytics:hours:in:focus:blocks:v3:flow – daily time in blocks of 2+ hours free from meetings, Slack, and email. Our current best metric for use in investigating a wider set of possible disruptions to deep work. This version is similar to v2, except with several algorithm refinements.

Zoom Attendance

Note that Worklytics only considers call attendance for users who are logged in to Zoom. If individuals are not required to be logged in to join internal Zoom calls (by organizational policy), and they join any calls without logging in to Zoom, their attendance in these calls will not be attributed to them in Worklytics' data. Check with your organization's Zoom administrator to ensure that all users are required to log in to Zoom for internal calls.

In order to prevent outliers from skewing the data, Zoom calls longer than 7h are capped at 7h. Typically, these are the result of rare cases where calls are left open for an extended period of time.

Welcome to the Data Dictionary. Here you will find information on how to leverage the Worklytics Datastream datasets for common analyses. Key resources are as follows:

The following is a sample set of SQL queries you can use to advance beyond the basics in your use of the Worklytics dataset. This code block assumes the user has access to the employee events dataset, though the exact table names will have to be substituted for the ones in your company's environment.

If you only have access to the individual aggregates, individual groups, and collaboration graph datasets, we recommend focusing your efforts .

If you're interested in understanding whether your team is in too many meetings, you're not alone. Meeting effectiveness is one of our most popular analysis topics.

To make it as easy as possible to dig into this area, here are a few metrics to start with:

1. How much time is our company spending in meetings?

Metric to consider: calendar:events:hours:meetings

Why this metric? More than 8 hours of meetings per week for knowledge workers tend to correlate with lower engagement scores, less focus time to accomplish deep work, and longer work days. It’s important to track and understand what portion of your organization is in this category in order to pinpoint where to take action.

2. How much of that time is in very large or recurring meetings?

Metric to consider: calendar:v1:meetings:attendees:10-or-more:hours AND calendar:v1:recurring:meetings:hours

Why this metric? Large meetings tend to be less productive and more conducive to distraction than do smaller group sessions. Similarly, recurring meetings typically stay on the calendar long beyond their usefulness, which is related to inefficiency and can bog down calendars. If you see high levels of either of these meeting types in your analysis, this is a good place to focus any interventions.

3. Are employees multi-tasking during meetings?

Metric to consider: calendar:events:focus

Why this metric? Distractions in meetings—especially large meetings—can be prevalent, and it's possible to quantify how much of meeting time is spent doing other types of work. While some usage of google docs and the like may be required for the meeting to be productive, having a value for calendar:events:focus below 85% is a red flag that distractions may be limiting meeting effectiveness.

4. How often do meetings start and end late?

Metric to consider: zoom:v2:late:meetings AND zoom:v2:overtime:meetings

Why this metric? Cleaning up meeting hygiene is often a quick win for teams that frequently start or end their meetings late. This, too, is quantifiable at the team level, and versions of this metric exist for both zoom and google meet.

5. Who is creating all the meetings?

Metric to consider: calendar:events:created

Why this metric? We tend to find in our analyses that close to 5% of people at a company create 60% of meetings, and for that reason, meeting related interventions are often best targeted at those people who have the greatest impact on meeting hours for the broader company.

5. How often do managers meet 1:1 with their direct reports?

Metric to consider: calendar:manager1on1:count

Why this metric? On a more positive note, some meetings tend to be associated with favorable survey outcomes, and one of these is manager 1:1s. When employees meet once every 1-2 weeks with their manager, they tend to report higher levels of engagement and more favorable perceptions of support. If you see pockets where 1:1s are happening infrequently, this represents an opportunity to drive engagement.

6. How often do managers co-attend their team's meetings?

Metric to consider: calendar:v1:manager-co-attendance:pct

Why this metric? While some amount of co-attendance can be good for coaching and training purposes or to move projects forward, we typically recommend managers attend about 20-30% of their direct reports' calls, and only in cases where their presence adds unique value. If you see very high or very low values for this metric, it's a sign that further investigation into management practices may be needed.

Below, you will find answers to our most frequently asked questions related to the dataset.

Dataset Format

Q: What do the dates mean?

A: Worklytics data is grouped by week. We consider weeks in ISO format, so weeks start on Monday and comprise Monday to the following Sunday (inclusive).

Q: What is a GROUP_TYPE in the context of this document?

GROUP_TYPE is any of: cost center, business segment, business unit, cost center, department, level, other affiliations, region, territory, role or the special maingroup, which is equivalent to the primary group type chosen by the customer.

Q: What does CUSTOM_GROUP_X mean in the context of this document?

Custom groups represent any additional HR data fields provided by your company outside of the standard fields (e.g., department, level, role, etc.). These will vary by organization.

Q: What does cv12wk mean in the context of this document?

Values with the suffix "cv12wk" are the "coefficient of variation" for the underlying variable over the trailing 12 weeks. Coefficient of variation is the standard deviation of the variable divided by its mean. We use the coefficient of variation because, unlike standard deviation, it does not depend on the unit of the variable. Eg, you can sensibly compare cv12wk for a variable measured in minutes with one measured in hours. We colloquially describe this as "volatility" because the greater the typical change in the underlying variable from week-to-week, the higher this number will be. So if one variable has cv12wk of 1, and the other has cv12wk of 2, the other can be thought of as "twice as volatile".

Q: How are meeting time values expressed?

A: Meeting time is expressed as hours in decimal format (e.g. a value of 1.25 equates to 1 h 15 min). You can multiply by 60 to get the amount of time in minutes.

Q: How does metric versioning work?

A: In certain cases, you may see metrics that appear similar but are differentiated by a v1, v2, or v3. In this case, the highest number represents the most up-to-date metric, though there are edge cases where you may want to use an earlier version. If no version can be found in the metric name, this is considered a v0 metric (earliest version).

Metric Definitions

Q: How do the versions of focus time differ? Which one should I use?

A: In most cases, v3:flow, as this takes interruptions from meetings, email, and chat into account and reflects our most up-to-date methodologies (v3:flow is based on a 2-hour block). However, certain roles may be better suited to the v3:prep or v3:focus versions of the metric.

To allow for further understanding of focus time and the related areas of opportunity, we also define and measure fragmented and interrupted time.

Essentially, fragments are blocks of time too short to be considered focus time but otherwise uninterrupted. And interrupted time adds up all the fragments and time spent on activities considered to be interruptions in order to calculate how much of the day is interrupted.

Q: What is a collaborator? A strong collaborator?

A: A collaborator is anyone you’ve interacted with through email, slack, meetings, or other collaboration tools in a given week. This is an estimate of your weekly work network.

A strong collaborator is the subset of these people with whom you’ve spent 2+ hours that week. This is an estimate of your day-to-day working group.

Q: What is the difference between inter- and intra-department collaboration?

A: Inter-department collaboration include at least one person from another team. Intra-department collaboration are single-team. These metrics can be calculated at the team, department, or other level.

Q: Are there thresholds in place for what is considered "collaboration"? Are meetings above the threshold completely removed from the dataset?

A: Unless your company has set a different threshold, events larger than 75 are excluded from collaboration-related calculations (e.g., we don't consider someone to have 10,000 collaborators because they attended an all-hands meeting). However, that very large meeting would still appear as a meeting in the dataset, given that it is an event that took up employee time.

Q: When might collaboration totals between a pair differ?

Meetings are a bidirectional form of collaboration, though slack and email (for example) occur in a single direction. The example below illustrates a case where collaboration is asymmetric between two parties over the course of a month.

Q: Why are collaboration time aggregate values greater than the total time in a given week in some cases?

A: For collaboration time aggregates (E.g., GROUP_TYPE_individual_collaboration_interteam_hours), this can happen because these metrics are a sum of the total time spent collaborating with each individual associated with a particular work item. For instance, spending 1 hour in a meeting with 4 team members will be counted as 4 hours collaborating with your team.

Working with the Data

Q: Which aggregates are used most frequently by customers?

A: Our Top 10 most popular aggregates are:

See List of Aggregates in the Data Export documentation for the full list of aggregates.

If you've worked with our team on custom reports, you will likely be familiar with the jitter plot visualization. This is a favorite way of ours to visualize Worklytics data because it is relatively easy to interpret, while also providing a granular level of detail around outliers and variance. See an example below:

To create these charts, we use the {Weekly_Individual_Aggregates} dataset.

First, we take a person average for each individual that meets the inclusion criteria for the analysis, which helps reduce the impact of outliers on the end visualization. The blue dots in the visual above represent each individual's average for the time period in question.

From there, we calculate the median value within a given group among the person averages. For example, the median Product employee in the example above has 22 collaborators on average.

Note that this visualization method can be used for any metric.

We recently published a blog about the concept of manager facetime, which we believe is important regardless of whether you work at a remote, hybrid, or fully in-person org. Specifically, employees in our analysis that saw their manager in-person more recently reported higher levels of satisfaction with the amount of coaching and support that their manager provides.

If you're interested in seeing what this looks like for your organization, here's how:

Requirements

• You'll need access to event-level data ({Employee_Events})

• An active badge data connector

• Accurate manager-to-direct report mappings

Methodology

The overall concept starts by determining the dates on which both members of a manager-direct report pair badge into the same office on the same day.

Tactically speaking, your query to identify all instances of badges into a building will look something like this:

SELECT week,
       person_id,
       time_first_visit,
       building_id
FROM (
       SELECT week,
              ARRAY(SELECT props FROM UNNEST(eventProperties) AS props WHERE props.name='buildingId')[OFFSET(0)].value AS building_id,
              employeeId AS person_id,
              MIN(eventTime) AS time_first_visit
       FROM {employee_events}
       WHERE 1=1
       AND week >= '2024-01-01' AND week < '2024-03-31'
       AND sourceKindId ='badge'
       AND employeeId IS NOT NULL
       GROUP BY 1,2,3)
;

From there, you can identify the instances where badges in by the employee and manager happened in the same building on the same day (see query 4.1 here for a similar, tactical example).

Next, to determine how long it's been since facetime has been possible, we calculate the difference in weeks between the current date and the date when we last saw an instance of facetime.

Finally, we aggregate this metric as described here to generate the final visualization.

Notes & Caveats

Is it possible to badge into the same office as your manager and still not see them? Of course. For that reason, we recommend thinking about this metric as an upper bound on how recently facetime could have occurred. In reality, it may be less frequent than badge-ins suggest.

It’s also possible to mitigate false positives by adding an analysis step to ensure a meeting between the manager and the individual contributor happened on the day in question. And team-wide offsites or known company gatherings can be incorporated on an ad hoc basis into this calculation.

Worklytics offers high-level guideline ranges to all customers and has a formal benchmarking dataset that is available to participating companies.

Overview of Guideline Ranges

Worklytics establishes guidelines for work activity metrics with the goal of providing organizations with recommended ranges for common values. The guidelines provided are based on industry best practices and the direct experience of the Worklytics team. They are not an industry benchmark.

The Guidelines are intended to reflect an estimate of the ideal range of activity for a typical knowledge worker. These types of employees would be expected to have ample time for deep thinking and focused work, as well as be less likely to experience burnout. Ranges are based on average weekly values per person over a 12 week period.

Here is a sample of guideline ranges for frequently-used metrics:

Overview of Benchmarks

Benchmarks offer more specific insight into how your company's work patterns compare to those of other companies. Benchmarks are available to participating companies and represent tens-of-millions of records and tens-of-thousands of employees across six different industries.

We chose our analysis approach to offer maximal clarity, while respecting a diverse set of participating firms' privacy, security and data protection expectations. Ultimately, we estimate benchmarks using Monte Carlo simulation and report fifth-quantiles for a subset of metrics at the manager, IC, and all-employee level. Future versions of the benchmarks product will incorporate additional cuts of data, among other improvements.

See the illustrative example below.

Note: If you are interested in exploring benchmarks for your organization and are not already a participating company, please contact info@worklytics.co.

When to use each

We typically use the guidelines to highlight "what good looks like" and how that compares to the reality at a company.

Benchmarks can be more specific and offer a point of comparison between your company and other organizations. However, as we see with focus time (example from our latest Industry Report below), most employees do not get the ideal amount on a regular basis. At a glance, having an org-wide value close to the 50th percentile of our focus time benchmark may seem like a good thing, but it would actually suggest your company needs more focus time if compared to the guideline range. We therefore recommend using the guideline ranges and benchmarks in combination with each other.