This guide is designed to help you assess the health of your employee data and identify areas where your data may need improvement. Retention Insights includes Data Health tools that will automatically scan your employee data and notify you of most potential issues.

If we discover potential issues in your employee data, we’ll display a banner at the top of your Retention Insights dashboard to notify you:

Clicking the View Data Health button will open the Data Health page, where you can learn more about the issues in detail.

If you don’t want to be warned about data health issues, you can click the Dismiss button to hide the banner. As long as there are unresolved issues found in your data, you’ll see a yellow warning icon next to the Data Sources button in the sidebar:

You can access the Data Health page by clicking the View data health button from the health check banner, or by clicking the Data Sources button in the sidebar and selecting the Data Health tab.

The Data Health page displays a list of possible issues that we’ve detected in your employee data. Each section can be expanded to provide a description of the issue and steps you can take to resolve it.

To resolve a specific data health issue, read the guidance in this support document and follow the instructions for your relevant HRIS integration, or for using a CSV import if you manually sync your employee data with Culture Amp.

Please note that retention data is refreshed once a day, so it may be up to 24 hours before your changes are reflected in the platform. You can see when your data health was last checked at the top of the Data Health page:

If you’re unable to resolve a particular issue or you don’t believe that it’s having a significant impact on your retention data, you can ignore it by toggling the Ignore switch:

When an issue is ignored, we won’t warn you about it in the Data Health Check banner. Issues will remain ignored until you manually toggle the switch again. Issues are only ignored for your own personal account; other backend users in your organization will still be shown the warnings unless they choose to ignore them themselves.

This displays the total number of employee records that have been detected as duplicates and automatically merged for your Retention Insights dashboard. For an understanding of how and why we automatically merge duplicate records, please refer to Automatic removal of duplicates.

If employee records are being incorrectly de-duplicated, this can result in a lower number of employees appearing in your dashboard. If you see a significant number of duplicate records, we recommend you check the Size & Composition chart to ensure that the headcount numbers align with the number of employees in your organization.

If you are seeing lower headcounts than you would expect, please refer to Automatic removal of duplicates to understand how our de-duplication process works and how you can update your data to minimise the risk of false positives.

This displays the total number of employee records with a missing or bad start date, including:

Any employee records without a valid start date are ignored by Retention Insights, which may result in lower headcount figures and inaccurate retention calculations.

If you are being warned of significant numbers of employees with bad start dates, we strongly recommend you ensure that all employee records including past employees are updated to include a valid start date.

This notifies you of large numbers of employees sharing an exit date. When we detect that a significant number of employees are sharing an exit date (more than 10% of your total headcount at time of exit), an additional notification is displayed above charts in the Overview section:

Shared exit dates will result in spikes in the Exited data in your Size & Composition chart, as well as downward spikes in your retention rate.

While there are normal circumstances in which a large number of employees exit your organization on the same date, this can also be a potential indicator of inaccurate exit dates in your employee data. The most common cause of inaccurate exit dates is using a partial import to update your employee data and not including terminated employees and their exit dates - see Include terminated employees in your employee data

If you’re warned about exit dates, we recommend you review the Size & Composition chart and assess whether the spikes in exits for the dates listed are accurate, or whether they may be a result of inaccurate exit dates. If the data looks accurate, you can safely use the Ignore switch to hide the warnings. If you think the data might be wrong, we suggest you update your employee data to accurately reflect each employee’s actual exit date.

To understand how you can minimize the risk of inaccurate exit dates going forward, please refer to Best practices for Retention Insights data.

This displays the total number of employee records that have a Name value consisting of a single word, ie, “John” and not “John Citizen”.

This displays the number of employee records that are missing an employee ID. While this won’t directly impact the data in your Retention Insights dashboard, it will increase the risk of inadvertently creating duplicate employees in the future - see Prevent duplicates for details.

We strongly recommend you provide unique employee IDs that cannot change over time when uploading your employee data, in order to reduce the risk of creating duplicate records.

This displays the number of employee records that are using an email as their employee ID. While this won’t directly impact the data in your Retention Insights dashboard, it will increase the risk of inadvertently creating duplicate employees in the future - see Prevent duplicates for details.

We strongly recommend you provide unique employee IDs that cannot change over time. Emails may be unique, but they can change in the event of an employee changing their name, your organization renaming or rebranding, etc.