Uploading Post-Hire Data

November 20, 2025 11:05

Harver helps you hire the best employees for the job. Depending on the model you are using, we help you hire employees who tend to stay with the company longer (optimized for employee retention) or we help you hire the employees that will perform best (optimized for employee performance).

To demonstrate the model's performance and the Harver assessments' predictive capabilities for employee performance and attrition, we require post-hire data. We've added a data upload section under the Insights tab to simplify sharing post-hire data.

Here you can read more about how the data upload works.

Required data

The full list of the data you can upload is available in the Data Checklist after clicking the “start” button in the platform. Notably, required data includes:

Data required to calculate attrition: Hired date (mm/dd/yyyy), Termination date (mm/dd/yyyy), Employee ID
Data required to match employee records to Harver assessment results: Email, First Name, Last Name

Additional data can be uploaded to increase the value of the analysis:

Attrition Reason, Voluntary/Involuntary Attrition
Grouping variables, e.g. Region, Department, Job Title

Step-by-Step instructions

Step 1: Upload a file

Head to Attrition data in the sidebar on the left, and select the 'upload file' option. Here you can upload your data as a CSV file.

If there is already some data in the Attrition data sheet, the ‘upload file' button won’t appear. You can still drag and drop a file into the page, or you can head to the Files tab in the sidebar.

This page shows a list of all the files you have previously uploaded. To upload new data, you first upload your new file, then click the “import” button that appears next to the filename.

The next page asks you where you would like to import your data to. Select the Attrition Data sheet.

Step 2: Map the columns in your file to the Harver template

You'll be redirected to a page where you can map each column from your file (shown on the left) to a column in the Harver template (shown on the right). This will help us understand the data you've provided.

Required “destination fields” are marked with an asterisk. Make sure that you link one of your columns (i.e. each “incoming field”) to one of the required columns. Each “destination field” can be used only once.

Step 3: Optional: assign values for the Voluntary/Involuntary Termination column

If you selected the "voluntary/involuntary termination" column, the system will ask you to map each of the values in your file to either "voluntary" or "involuntary."

Tip: If you have data on termination reasons that are not classified as voluntary or involuntary, you can use this step to classify them easily. In your original file, duplicate the column with termination data and map this duplicate column to the Voluntary/Involuntary Termination in the previous step.

Step 4: Fix any data problems

If the data you uploaded has any problems, these will be highlighted in red as invalid values. It is important to fix as many invalid values as possible before submitting your data, as they strongly affect the accuracy of the results.

Examples of invalid values include: missing required data (such as email), incorrect date format, and inconsistencies in hiring and termination dates.

Missing optional values will be highlighted in yellow as warnings. These are less important to fix before submitting.

You can fix invalid issues directly by double-clicking on the cell.

Transform (AI option, bulk update using natural language)

The transform button allows you to bulk update multiple records using natural language prompts.

Examples:

Split values in a column across multiple columns:

The last name column should be split, the text before the comma is the last name, the text after the comma is the first name
Change date format:

The hired date field is in EU date format (dd-mm-yyyy), change it to US format (mm-dd-yyyy)

Step 5: Review Harver Candidate IDs

Once you've uploaded your data, the system will search for Harver assessment results for each employee in your dataset. The system will first search based on each employee's email. If a candidate with the same email cannot be found, then the system will search by first name and last name.

If assessment results are found, the employee's candidate ID will show up in the Harver Candidate ID column.

As you can see above, the Candidate ID column can have three other types of values:

Hired before Harver: the employee's Hired Date occurred before your Harver account was created.
Application still in progress: no completed assessments could be found. A specialist needs to review the in-progress assessment results to determine if they can be used for the study.
No match for email or name: the system couldn’t find any candidate results that match the employee's email or name.

Step 6: Submit data

Don’t forget to submit your data using the button in the top right corner! Until you submit, we won’t be able to analyze your data.

FAQ

Why don’t I see the ‘data upload’ page under the Insights tab?

Only users with Insight edit rights can access the data upload functionality. Reach out to your Harver account manager to request edit rights.

Why don’t I see the ‘upload file’ option under the Attrition Data tab?

The ‘upload file’ button is only available when you haven’t uploaded any data yet. If there is already some data in the Attrition Data Tab, you have two options:

Remove the data. Use the check box in the top left corner to select all records, then select ‘delete’ from the options that appear at the bottom of the screen.

Upload your data through the Files tab. This step requires more steps: first upload a file, then once it appears in the list, select “import”. In the next screen, select the “Attrition Data” option, and then continue to the mapping stage.

What type of files can I upload?

Currently, you can only upload your data as a CSV file.

What date format can I use?

We expect dates to be in the US format: mm-dd-yyyy or mm/dd/yyyy.

Why don’t all employee records have a candidate ID?

There could be several explanations for why we could not link one of your employee records to a Harver candidate:

The system could not identify the correct candidate. This can happen if the employee used a different email or a different name for the Harver assessment.

The candidate results were archived. The availability of Harver assessment data depends on the data retention rules you have set up in your Harver account. Once assessment data is archived, we can no longer identify which employee it corresponds to.

The employee never completed a Harver assessment.

Where does the ‘Candidate ID’ column come from?

The system adds the Candidate ID column after you first upload your data. The system searches through the Harver database to identify which candidate, i.e., which set of assessment results, corresponds to each employee record. This search is based first on email. If the system cannot find a candidate with the same email, it will search for the first and last name instead.

Why are there multiple Candidate IDs for some employee records?

In some cases, the same employee will have completed multiple Harver. For example, the recruitment process could have included primary and secondary assessments, or the employee could be a re-hire.

Can I go back and change the column mapping?

No, you cannot change the column mapping once the data has appeared in the Attrition Data sheet. We recommend deleting all the data in the Attrition Data sheet and importing the data again to change the column mapping.

How can I clear the Attrition Data sheet?

Select all the rows by clicking the checkbox in the top left corner of the sheet. A delete option will appear at the bottom of the screen.

In This Article: