> ## Documentation Index
> Fetch the complete documentation index at: https://docs.pollandroll.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Importing existing data

> Creating a survey from a file of data already collected, in Triple S format.

If your responses have already been collected elsewhere, you can import them straight into poll\&roll rather than starting over from a blank questionnaire. The survey created skips the **Questionnaire** and **Data collection** steps: once the import is finished, it is ready right away for the **Results** and **Report** tabs.

<Note>
  The expected format is **Triple S**, a standard exchange format in marketing research: a **metadata** file that describes the variables, and a **data** file that contains the responses.
</Note>

## Start the import

From the home page, click **Create a survey**, then choose **Import an existing data file into a dashboard**.

<Steps>
  <Step title="Name the survey">
    Give it a title.
  </Step>

  <Step title="Upload the metadata file">
    **.xml** or **.sss** format: it describes the variables present in the data file (name, label, type).
  </Step>

  <Step title="Upload the data file">
    Fixed-width **.dat** or **.asc** format, or comma-separated **.csv**.
  </Step>

  <Step title="Choose the owner organization">
    This is the organization that will own the survey created.
  </Step>
</Steps>

<Note>
  Both files are checked before you continue: expected format, reading of the metadata file, recognized Triple S version, and matching interview identifiers between the two files. An error blocks the step until it is fixed.
</Note>

## Select the variables to import

You then land on a table listing every variable in the metadata file (its name and its label); all are included by default. Tick one or more variables to:

* **Exclude**: leave them aside, they will not be imported.
* **Group**: bring them together into a single question with several items, as in the questionnaire editor — useful when one and the same grid question is split across several variables in the data file. Grouped variables are then highlighted in the same color.

Once your selection is made, click **Import**.

## Track the import

The import runs in the background; the survey shows the **Data processing** status in the meantime. You receive an **e-mail** once the import is finished, with the number of interviews imported and, where applicable, those that were not.

### Interviews not imported

An interview with a value that does not match what the metadata file describes (wrong type, decimal number written with a comma, value outside the declared bounds…) is **skipped individually**: it does not prevent the rest from being imported and is not counted in the results.

A banner on the survey **Summary** warns you if interviews were skipped. Click **See details** to find, for each one, its line in the file, its identifier and the reason it failed.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.