> AI agents: this is one page from Mammoth Analytics documentation. The index of all pages as Markdown is https://mammoth.io/docs/llms.txt. Append `.md` to any docs URL, or send `Accept: text/markdown`, to get Markdown.

# Send pipeline output to another dataset in Mammoth

> Send a View's cleaned rows to a new or existing Mammoth dataset with Send to Dataset. Map columns, keep rows up to date or add a copy on each run.
>
> Applies to: All plans. Roles: Project Editor or Project Admin. Where: View › Describe your transformation › Unify › Send to Dataset

**Send to Dataset** sends this View's processed rows to a new or existing dataset, for reporting or more work. Several Views can send to one dataset to stack their rows. Open it from **Describe your transformation**, then **Unify** › **Send to Dataset**.

## Send rows in three steps

1. **Choose where the rows go**, then click **Next**:
   - **New Dataset**: type a name. **Location** starts in this dataset's folder, or the place you last sent to; click the edit icon to pick another.
   - **Existing Dataset**: pick it from the list (search to find it).
   - **One per value**: pick a text column with 2 to 49 values, none longer than 100 characters, and a name for the datasets. The name must include the column token, so each dataset is named after its value. Each value gets its own dataset, in its own folder, inside a folder of this project.
2. **Map the columns and choose what each run does** (see the two tables below).
3. **Check where the task runs.** **Task Position** shows **End of Pipeline**, so the task stays last: tasks you add later go above it. To change it, click the edit icon and untick **Keep at the end of the Pipeline**. Then click **Apply**.

![Column mapping step with source columns on the left connected by arrows to destination columns](https://mammoth.io/docs/images/20260417_173748_Thecolumnmappingstepshowingsourcecolumnsontheleftconnectedbyarrowstodestinationcolumn1.webp)



## Column mapping options

| Option | What it does |
| --- | --- |
| **All columns** | Switches every column on or off at once. Then switch single columns back on or off. For a new dataset, every column starts on. |
| Column rows | Switch a row off to leave that column out. For a new dataset, the up and down arrows set the column order. For an existing dataset, pick the destination column from the list, or **Create a new column**. |
| **Hide already-mapped columns from selection menu** | Existing dataset only, on by default: the list leaves out columns already mapped. |
| **Include hidden columns** | Also sends columns hidden in this View. Shown only when the View has hidden columns. |
| **Add Condition** | Sends only the rows that meet a condition |

## Keep rows up to date or add a copy each run

| Each time this runs | What happens | Use it for |
| --- | --- | --- |
| **Keep this view's rows up to date** (default) | Replaces the rows this View sent last time. Rows from other Views stay. | A destination that always holds the latest output |
| **Add a new copy of all rows** | Keeps every earlier run, so the dataset and its totals grow | History, such as a weekly snapshot of sales |

When other Views also send to the dataset, a note names them, for example "Also fed by 2 other views: North sales, South sales." It lists up to four, then "and 3 more". With **Keep this view's rows up to date**, each View replaces only its own rows, so several Views can share a destination safely.

Mammoth remembers the location, update method and task position you last used, and starts there.

Switching an existing task from **Add a new copy of all rows** to **Keep this view's rows up to date** warns: "The next run also removes the copies this view added on earlier runs."

## Related

- [Pause sends while you edit](https://mammoth.io/docs/learn/clean-and-shape/pause-sends/): hold Send to Dataset while you change the pipeline.
- [Join two datasets on a shared column](https://mammoth.io/docs/learn/clean-and-shape/join/): add columns instead of rows.
- [Combine datasets](https://mammoth.io/docs/learn/organise/combine-datasets/): stack files in the Data Library.

---
Source: https://mammoth.io/docs/learn/clean-and-shape/send-to-dataset.md · Updated: 2026-11-08