Employee survey governance
Employee survey data segregation for multiple business units: A complete guide
Employee survey data segregation keeps each business unit’s results behind its own access wall. This guide covers setup, governance, group roll-ups and vendor questions.
Reviewed by Siva Samraj
Director of Engineering, CultureMonkeyLeads the engineering team behind the CultureMonkey platform
Connect on LinkedIn →“When survey data leaks from one business unit to another, it usually starts with something small, like an admin role set too wide on day one or a filter that narrows a small team down to a few people. That’s why the walls belong in the data layer, and every admin scope deserves a test before the survey goes live. Once employees trust that only their own entity’s HR team reads their comments, they write what they really think.”
Why does data segregation matter in a multi business unit employee survey?
Data segregation matters in a multi business unit employee survey because each subsidiary is its own employer, with its own legal duties, leaders and employee trust to protect. One shared survey program works only when every business unit’s results stay inside that business unit.
Legal entity boundaries
Each subsidiary is usually its own employer and data controller. GDPR Recital 48 says a group “may have a legitimate interest” in sharing employee data for internal administrative purposes, a test each transfer has to pass.
Honest answers
Employees answer candidly when only their own HR team and leaders read the results. A belief that the parent company reads every comment lowers participation in every entity.
Works councils
In Germany, “Staff questionnaires require the approval of the works council” under Section 94 of the Works Constitution Act. Agree in writing who will see results before you ask for that approval.
Breach exposure
IBM and the Ponemon Institute put the global average cost of a data breach at USD 4.99 million in the Cost of a Data Breach Report 2026. Narrow access limits what one compromised admin account can expose.
Data segregation also keeps comparisons fair. When each business unit sees its own results first, entity leaders act on their own data, and group HR compares entities on scores that every unit has already checked.
Takeaway Treat every business unit as its own walled survey audience first, then build group views on top of those walls.
How does survey data segregation work across subsidiaries?
Survey data segregation across subsidiaries works in four layers: an org hierarchy, an entity attribute on every employee record, data mapped in from the HRIS, and access rules that turn those attributes into walls. Each layer depends on the one before it.
- Layer 1
Org hierarchy
Reporting lines from the manager field decide which managers see which teams. Leaders see their whole downstream organization.
- Layer 2
Entity attribute
Every employee record carries a legal entity or business unit value. The walls are built from that value.
- Layer 3
HRIS mapping
A scheduled HRIS sync keeps entity, location and manager fields current, so joiners, movers and leavers land in the right entity.
- Layer 4
Access rules
Roles decide what someone can do, and scopes decide whose data they see. Entity-level access gives each entity HR lead a scope of one business unit.
Most groups, from a holding company with a few subsidiaries to a conglomerate with dozens, choose between one survey account with an entity attribute and a separate account for each subsidiary. One account keeps group roll-ups simple and needs careful scoping. Separate accounts give the hardest wall and make group comparisons harder to build.
Entity-level segregation decides which subsidiary’s data a person can reach. Role-level permissions, such as what a manager or HR business partner may open inside that subsidiary, are covered in this guide to role-based access control for engagement surveys. Large groups usually need enterprise survey software that supports both layers.
Takeaway Survey data segregation is only as accurate as the entity field on each employee record, so fix the HRIS data before you build the walls.
“CultureMonkey made it easy to automate our employee engagement surveys while still giving us the flexibility to customize them. Their real-time data analysis, combined with insights from the People Science team, helped us uncover key focus areas, align our strategy with industry benchmarks, and strengthen our leadership development efforts.”
- 80%+participation across 2,000+ employees
- +10%above the financial industry engagement benchmark
- 15business units with clear clarity
How do you set up separate survey dashboards per company or business unit?
Set up separate survey dashboards per company or business unit in seven steps, from mapping legal entities to testing each admin’s view. Group HR owns the design, IT owns the data feed, entity HR owns the checks, and the vendor configures the platform.
- 01
Map entities and reporting lines
List every legal entity, business unit and shared-services team, and decide which entity owns staff who work across several.
- 02
Add an entity field in the HRIS
Make the legal entity or business unit a required field so that no employee record arrives blank.
- 03
Sync the hierarchy to the survey platform
Connect the HRIS and check that the headcount for each entity in the platform matches the headcount in each subsidiary.
- 04
Set anonymity thresholds
Pick one minimum for the group, and raise it for small entities or where an agreement with employee representatives requires it.
- 05
Create scoped admins for each entity
Give each entity HR lead access to their own business unit only, and keep group-wide admins to a short named list.
- 06
Test every scope before launch
Preview the platform as each admin and confirm no other entity’s data appears. In some platforms a blank scope means access to everyone.
- 07
Launch and share entity dashboards
Each entity HR lead shares results with their own leaders, and group HR reviews the roll-up across entities.
Once the walls hold, each entity’s dashboard shows only that entity’s people, and the same setup carries over to every later employee engagement survey cycle.
Takeaway Testing each admin’s view before launch is the step that catches a broken wall, so never skip it.
What governance rules should a conglomerate set for employee engagement surveys?
A conglomerate should set written rules for who may create surveys, send them, see scores and comments, export data and keep it. Group HR writes the rules once, every entity follows the same rules, and Legal and IT check them.
| Action | Group HR (parent) | Entity HR (subsidiary) | Legal and IT |
|---|---|---|---|
| Create surveys | Owns the core question set | Adds a few local questions | Reviews any new data field |
| Send surveys | Sets the group survey calendar | Sends to its own employees | Approves channels and sender domains |
| See scores | Group roll-up and entity scores above the threshold | Own entity only | No routine access |
| See comments | Entity theme summaries only | Own entity, above the threshold | No routine access |
| Export data | Group summary only | Own entity, with exports logged | Approves any raw export |
| Keep data | Sets one retention period | Follows the group period | Checks it against local law |
| Set thresholds | Sets the group minimum | Can ask for a higher minimum | Confirms works council terms |
| Add an acquired company | Decides the first survey date | Runs the first survey | Updates the data agreement |
GDPR’s data minimization principle applies to these rules as well: personal data should be “limited to what is necessary” for the purpose. A parent that needs group trends rarely needs comment-level data from every entity.
These employee survey best practices cover the calendar, reminders and communication that sit around the governance rules.
Takeaway Write the governance rules into one table that every entity signs off, and review it every year and after every acquisition.
· S03 E07“I think the biggest mistake companies make with regard to culture is putting it all on the HR department and keeping it separate from how the business runs. Culture has to be integrated so it becomes part of who they are and how they work, and that is everybody’s responsibility.”
See each business unit’s results behind its own wall
Scoped admins for every entity, one anonymity floor for the group, and roll-ups for group HR.
How do you compare engagement across business units without breaking segregation?
Compare engagement across business units by rolling up scores instead of raw responses, and by measuring each unit against a fair baseline. Group HR sees entity scores side by side, while comments and small groups stay inside each entity.
Share one core: keep a set of core questions word for word in every entity, built from one employee engagement questionnaire, so a score means the same thing everywhere.
Roll up scores only: group views show averages and participation by entity, and comments stay with the entity HR team.
Hold the threshold at every level: an entity, location or team below the minimum shows no score at group level either.
Compare like with like: benchmark each business unit against its region and size, as well as against the group average.
Region moves the baseline before any entity effect
Engagement score on the 0 to 10 scale, shown from 7.0 to 8.5. The dark line marks the median across regions.
Tap a region to compare it with the regional median of 8.09.
In CultureMonkey’s own benchmark data, APAC and LATAM both average 8.22, while Europe averages 7.67 with an eNPS of +8. A European subsidiary that trails an Asian sister company may simply match its region.
Source: CultureMonkey People Science, Employee Engagement Benchmarks by Region 2026, Q3 2026 edition.
Takeaway Group-level views should compare entity scores against fair baselines and never pull raw comments across the wall.
“When we were just using forms, we were limited to just sentiment information and eNPS information across our portfolio of companies. We couldn’t really drill down into it beyond the company level.”
“We were having these conversations in the leadership team about how we’re actioning survey data. If we’re getting all this information every twice a year, we’re really not doing anything with it if we’re really honest and we look in the mirror.”
- 20+partner companies covered in one survey
- 2,400responses from 20+ partner companies
- 7.5engagement score, first structured baseline
- 6first eNPS baseline
What are common mistakes when segregating survey data across entities?
The most common mistakes when segregating survey data across entities come from treating a filter as a wall, setting thresholds too low, and forgetting entities that change. Each mistake can be caught before launch.
- 1
Confusing segmentation with segregation
A report filter that an admin can remove leaves every entity visible. Put the wall on the admin account itself.
- 2
Setting thresholds too low for small entities
A 12-person subsidiary with a threshold of 3 can expose individuals once results are cut by team. Raise the minimum for small units.
- 3
Adding too many group admins
Every extra group-wide admin is one more person who can see every entity. Review the admin list every quarter.
- 4
Copying one entity’s questions to all
Questions written for a head office may not fit a plant or a retail brand. Keep a shared core, let each entity add a few local questions, and run multilingual surveys where entities work in different languages.
- 5
Ignoring acquired companies
A newly acquired company often has no entity field yet, so its people land in the wrong scope or in none. Map the company before its first survey.
- 6
Judging small subsidiaries against the group average
A 60-person subsidiary and a parent with 5,000 people start from different baselines. Compare each entity with companies of its own size first.
Company size moves the baseline too
- 8.16
- Enterprise (2000+)
- 8.09
- SMB (100-500)
- 8.07
- Mid-Market (500-2000)
- 7.70
- Micro (<100)
In CultureMonkey’s own benchmark data, companies with fewer than 100 employees average 7.70 on the 0 to 10 scale, while companies with 2,000 or more average 8.16. Compare a small subsidiary with companies of its own size before reading its gap to the group.
Source: CultureMonkey People Science, Employee Engagement Benchmarks by Company Size 2026, Q3 2026 edition.
Takeaway Most segregation failures are setup failures, so review scopes, thresholds and entity fields before every survey cycle.
What should you ask a survey vendor about multi-entity data segregation?
Ask a survey vendor these 11 questions about multi-entity data segregation before you sign, and ask for each answer in writing. They cover walls, thresholds, roll-ups, data location and what happens when an entity leaves the group.
Access walls
- 01
Can an HR admin in one subsidiary see another subsidiary’s results?
Why it matters One wrong scope exposes a whole entity’s scores and comments.
Good answer Scope is enforced on the data, and the admin cannot remove it.
- 02
What does an admin with an empty scope see?
Why it matters Some platforms treat a blank scope as access to everyone.
Good answer The vendor names the default and shows how to test it.
- 06
Does any comparison score use data outside the viewer’s scope?
Why it matters An organization-wide average can show group numbers to an entity admin.
Good answer The vendor lists which benchmarks ignore scope, and why.
Anonymity
- 03
Does the anonymity threshold apply to every filter, export and roll-up?
Why it matters A threshold that skips exports or roll-ups leaks small groups.
Good answer One minimum applies on screen, in exports and in group views.
- 04
Can each entity have its own anonymity threshold?
Why it matters A works council agreement may set a higher minimum in one country.
Good answer A clear yes, or a documented way to raise it per survey.
Roll-ups and structure
- 05
Can group HR see a roll-up without seeing entity comments?
Why it matters Scores and comments carry different risk at the parent level.
Good answer Roll-up access and comment access are separate permissions.
- 07
Can each subsidiary run in a separate workspace, and can results still roll up?
Why it matters Separate workspaces give the hardest wall and complicate group views.
Good answer A written description of both models and how roll-ups work in each.
- 08
Which HRIS fields set entity membership, and what happens when someone moves entity?
Why it matters Movers can drag past answers into the wrong entity.
Good answer Past responses keep the attributes they had when submitted.
Data and exit
- 09
Where is each entity’s data hosted, and can entities use different regions?
Why it matters Data residency rules can differ between subsidiaries.
Good answer Named hosting regions and a data processing agreement.
- 10
Which sub-processors handle open text comments, including any AI analysis?
Why it matters Comments hold the most identifying content in a survey.
Good answer A current, published sub-processor list.
- 11
How is one entity’s data exported or deleted after a divestiture?
Why it matters A sold business unit takes its employees, and may need its data.
Good answer Per-entity export and a written deletion process.
Replacing a tool? This guide to switching employee engagement survey vendors covers keeping score history, and this comparison of employee engagement survey software lines up the main platforms.
Takeaway A vendor that answers every question in writing, including the ones it cannot support yet, is a safer choice than one that answers only in a demo.
How does CultureMonkey support data segregation for multiple business units?
CultureMonkey supports data segregation for multiple business units with business groups and units, admins scoped to one entity, an anonymity floor that holds on every filter, and a read-only HRIS sync. Each part below is documented in the CultureMonkey Help Center.
Business groups and units
Map each subsidiary or legal entity to a business unit or business group, then filter or compare one entity against another. Business groups are turned on with an account setting.
Help Center: Business groups and unitsSub-admins scoped to one entity
Scope each entity HR lead by business unit, business group, location or team. The scope applies at the data layer, so it cannot be switched off from a report.
Help Center: Sub-admin data scopeAn anonymity floor of 3
Anonymous surveys never report below 3 responses, and the minimum can be raised per survey. Business group and business unit filters respect the same threshold.
Help Center: Response thresholdRead-only HRIS sync
A one-way, read-only sync with HR systems such as Workday and BambooHR brings in new joiners, flows team and manager changes through, and deactivates leavers.
Help Center: Connect an HRIS
The CultureMonkey security and compliance page lists ISO 27001:2022 certification, a SOC 2 Type II audit, and separate US and EU environments. For separate accounts per subsidiary or roll-ups across accounts, put the vendor questions above to the CultureMonkey team as well.
Conclusion
Employee survey data segregation for multiple business units keeps each entity’s survey responses, dashboards and exports inside that entity, while group HR still sees a fair roll-up. Data segregation is the governance layer that makes one survey program safe across subsidiaries, brands and acquired companies.
This guide covered why segregation matters, the four layers that build entity walls, a seven-step setup, a governance rules table, fair comparisons across business units, common mistakes, and 11 questions to ask every vendor.
CultureMonkey serves that goal with business groups and units, sub-admins scoped to one entity at the data layer, an anonymity floor of 3 on every business group and business unit filter, and a read-only HRIS sync that keeps the employee list current.
Frequently asked questions about employee survey data segregation
What is employee survey data segregation?
Employee survey data segregation is the practice of walling each business unit’s or subsidiary’s survey responses, dashboards and exports so only the people responsible for that entity can see them. Data segmentation only slices results by team or location to find patterns.
Can HR in one subsidiary see another subsidiary’s survey results?
HR in one subsidiary should not see another subsidiary’s survey results. Give each entity HR lead an admin scope limited to their own business unit, and test the scope before launch, because in some platforms a blank scope means access to everyone.
Should a parent company have access to subsidiary survey data?
A parent company should see subsidiary survey data as roll-up scores and participation above the anonymity threshold. Raw comments and small-group results should stay with the subsidiary’s own HR team unless a written policy says otherwise.
How do you run engagement surveys after a merger or acquisition?
After a merger or acquisition, map the acquired company as its own business unit, run a short baseline with an eNPS survey template, and add it to the group calendar from the next cycle. Teams leaving another tool can follow these survey migration guides.
Do works councils need to approve employee engagement surveys?
In Germany, works councils must approve staff questionnaires, and other countries have their own consultation rules. Share who will see results before you ask, as this guide to anonymity for frontline and unionized employees explains.
What is the minimum team size for reporting across entities?
Set a minimum of at least 3 responses for any group, and raise it for small subsidiaries where team cuts could point to individuals. Apply the same minimum to every filter, export and roll-up, as any anonymous employee feedback tool should.
Does data segregation slow down survey reporting?
Data segregation adds setup time, mainly mapping entities and testing scopes, and no reporting delay afterward. Each entity HR lead opens their own dashboard once results clear the threshold, and a pulse survey tool reuses the same walls every cycle.
How does GDPR apply to multi-entity employee surveys?
Under GDPR, each subsidiary is usually a separate controller, so sharing survey data with a parent needs a lawful basis and a clear purpose. Article 88 adds employment rules on transfers within a group. Collect only what each entity needs, and record who may see it.
Can each business unit add its own survey questions?
Each business unit can add a few local questions on top of a shared core that stays word for word the same across the group. Keep local questions short and specific, such as a block of manager feedback survey questions for a unit with new team leads.
How should a subsidiary respond to negative survey comments?
A subsidiary should respond to negative survey comments inside its own entity, with its own leaders, and share themes rather than quotes with group HR. These negative feedback examples show how to respond without getting defensive.
Run one survey program with every entity’s data kept separate
Business units, scoped admins and group roll-ups, set up with the CultureMonkey team.