Skip to main content
Raw Data Import lets you bring existing feedback — for example from a previous survey tool — into Honestly by uploading a CSV file. Each row becomes one anonymous response in a survey you choose. Open Imports ↗ · Opens the webapp — please log in first · Requires the Raw Data Import feature · Admins only
The import runs in the background. After you upload, the file is queued and checked while you keep working — the result appears in Your imports a moment later, not instantly.

​
1. Set up the target survey

Before you import, the survey you import into must already contain everything the file refers to:
  • Create the survey with the same questions and answer options used in the tool you are importing from.
  • The question wording and each answer option must match your file exactly — the import matches on the text, not on a position or an ID.
  • Create the employee attributes you want to import, for example Department. You do not need to create the attribute values themselves — those are created automatically from the file.
Two of the survey’s own question states catch people out here:
  • The Hide question toggle only affects the questionnaire participants fill in — the import does not read it. A hidden question whose text is used as a column header is filled like any other, and those responses appear in your reports even though the question is shown to no one. Hiding is also not a way around a repeated wording: if four questions share the same text and three of them are hidden, the import still finds four matches and fails before writing a single row.
  • Delete question removes a question from the import entirely. A deleted question no longer counts when the import checks a wording, but it can no longer be filled either — a column carrying its text fails the whole file with Question not found.
A question that has no column in your file is not an error: the import reads the file’s columns, not the survey, and simply leaves that question empty.
When you check the survey for a repeated question wording, count the greyed-out hidden questions too — the import counts them.

​
2. Prepare your CSV file

Download the Excel template from the import page and fill it in, then save it as CSV (comma-separated, not semicolon). The header row decides how each column is read:
  • The first column must be named Date and hold each response’s date.
  • Add each survey question as its own column, with the header matching the survey wording exactly.
  • Add at least one attribute column, prefixed with attribute: — for example attribute:Department. At least one attribute column is required; without one the import fails with No attributes found in import file.
Dates use the format DD.MM.YY (for example 07.07.25). You can optionally add a time: 07.07.25 14:30 or 07.07.25 14:30:00. This applies to the Date column and to the cycle columns described below. Fill in each answer the way its question type expects:
  • Single select, multiple select and ranked questions — enter the answer option text, exactly as it reads in the survey. Not a number. For a multiple select question, put several options in one cell separated by commas.
  • Smiley questions — enter the number of the scale point: 1 is the best, higher numbers are worse.
  • Star questions — enter the number of stars: 5 is the best, 1 is the worst.
Answer options are matched in the language of the person who starts the import. Enter the answer texts in that person’s account language.
Attribute values are created automatically as they appear in the file, up to 100 characters each.

​
3. Import across several survey cycles

By default a whole file lands in a single cycle spanning from its earliest to its latest date. To split one file across several cycles — for example twelve months of a monthly survey in one upload — add two optional columns, CycleStart and CycleEnd. Each row then declares which cycle it belongs to. The rules are strict, and breaking any of them stops the whole import:
  • Include both columns or neither. A file with only one of them is rejected.
  • Every row must fill in both values.
  • CycleEnd must be after CycleStart.
  • Each row’s Date must fall inside that row’s own cycle.
Rows that share the same start and end belong to the same cycle; each distinct pair creates its own cycle. If a cycle with exactly those two dates already exists, it is reused — so you can add more responses to an existing cycle later. Cycle dates use the same format as the Date column. A CycleEnd given as a date only covers the whole day through 23:59:59. Without these columns the original behaviour stays: one cycle from the earliest to the latest date in the file — or, if a cycle already spans that period, your responses fall into it.
Imported cycles have no invitation data, so their participation rate stays empty.

​
4. Run the import

  • Open Imports.
  • Click Create new import, then Create import.
  • Under Select survey, search for and choose the survey to import into, and double-check it is the right one.
  • Under Select file, choose your prepared CSV. Keep it below the size limit.
  • The upload only checks that a non-empty file arrived. The full check of the contents happens afterwards, in the background.
Watch the row in Your imports for the Status: Processing…, then Imported or Import has errors.

​
5. If a row or the file has errors

There are two different outcomes, and it matters which one you get:
  • A structural problem stops the whole file. Nothing is imported and the status becomes Import has errors. This covers an unreadable date, a question or attribute column that does not match the survey, an incomplete CycleStart/CycleEnd pair, a row whose date is outside its declared cycle, or a file with no answers, no questions or no attribute column.
  • An unknown answer option skips only that one answer. The rest of the row and the rest of the file are still imported, and the status stays Imported. The skipped answer is listed in the error report.
Open the Errors report from the import row to see what was skipped or why the file failed.
Check the error report even on a successful import — a file can finish as Imported while individual answers were quietly skipped.

​
6. After the import

  • Imported responses appear in your reports and all standard analytics.
  • The cycles you declared (or the single fallback cycle) are available in the cycle picker, shown by their date range.
Deleting an import removes its responses. Cycles created from CycleStart / CycleEnd are left behind empty — only the single-cycle case is cleaned up automatically.