Skip to content

Automatic translation — the Portuguese version is the reference. View in Portuguese

Updated on 2026-10-04

Add a column to a Silver or Gold table

How to bring a new data field (e.g. clicks) into an existing datalake table — edit the recipe, run the pipeline and refresh the dashboard.

On this page (4)

The Silver and Gold tables (the mart_… ones too) are built by a recipe: which tables the data comes from, which columns are included and which filter applies. A new field — for example, clicks in a table that only has impressions — only appears in the table when it is added to the recipe and the table is rebuilt.

Table column × calculated column

If what you want is a calculation using fields the dashboard already has (average ticket = revenue ÷ orders), you don't need to touch the table: create a Coluna calculada in the dashboard editor. This tutorial is for when the data does not yet exist in the table.

Before: does the recipe already have the column?#

Sometimes the recipe already computes the column and the table was generated before that — in that case you only need to rebuild the table (step 4). To check, ask Aiô ("does mart_estado_cidade compute clicks?"): Aiô reads the recipe and the columns the table currently has, and tells you whether you only need to run it again.

Step-by-step#

  1. In ETL/ELT → Pipelines, open the pipeline that builds the table. Gold tables live in their own pipeline, separate from the one that brings data from the source.

  2. In the Camada Gold (or Camada Silver) section, find the card for the table and click Editar.

  3. Add the column:

    • in Guiado mode, use Adicionar coluna and choose the field from the source table — or Coluna calculada, for a calculation (e.g.: valor_total * 0.1);

    • in SQL livre mode, add the column in the final SELECT. Example, counting clicks from each source row:

      sql
      COUNTIF(Clicked) AS clicks
  4. Save. The card will show ⟳ requer reprocessar: the recipe changed, but the table hasn't been rebuilt yet.

  5. On the table card, click Processar agora — this runs only that table, for real, and shows the result with the Ver execução link. (To run the entire pipeline, use Executar agora at the top.) When the run finishes, the table will have the new column and the card will return to ✓ Processado.

  6. In the dashboard that uses the table, click Atualizar agora. This recreates the copy of the table the dashboard keeps, and the column appears in the list of Campos.

  7. Drag the column into the visual — numbers go to Valores; text and date go to Eixo (on maps, Local).

You will see the column in the Campos list under the table name, and the value in the visual.

Why not "bring the field from another table"?#

A field with the right name may exist in another table — clicks might be in the daily table, for example. But a table by day and another by state and city count things at different levels: summed together they produce a number that doesn't mean anything. The correct approach is to make the column originate in the table itself, at the same level of detail as the table (step 3).

Frequently asked questions#

I ran it and the column didn't appear in the dashboard. You're missing step 6: the dashboard reads a stored copy of the table, and Atualizar agora is what brings the new columns.

The card says "⟳ requer reprocessar" for days. The recipe was edited and the pipeline hasn't run since. Execute it (step 5) — or enable the pipeline schedule so it runs automatically.

Related links