How to Get Started with Sigma Computing

Sigma Computing gives non-technical analysts direct access to warehouse data through a spreadsheet-style interface. Instead of exporting data to Excel or waiting for an engineer to run a query, you open a Sigma workbook, connect to your Snowflake or BigQuery table, and start slicing rows the same way you would in a spreadsheet. Data stays in the warehouse. Nothing is downloaded.

Sigma went public in March 2025 and has grown at roughly 100% year-over-year for four consecutive years. As of early 2026, more than 1,000 companies use it as their primary BI layer, and it now ranks among the top-rated analytics platforms on G2 for ease of use by non-technical users.

What You Need Before You Start

You need three things: a Sigma account (free trial available), a supported cloud data warehouse, and credentials to connect them.

Sigma supports Snowflake, BigQuery, Redshift, Databricks, PostgreSQL, AlloyDB, and MySQL. If you already have a Snowflake or BigQuery account, you can connect in under five minutes. If you do not have a warehouse yet, Snowflake offers a 30-day free trial with $400 in credits, which is more than enough to run the examples in this guide.

A Sigma Organization Admin role is required to create connections. If you are setting up for a team, one person handles the warehouse connection and then invites teammates with viewer or explorer roles.

Step 1: Create a Sigma Account

Go to sigmacomputing.com and click Start Free Trial. Enter your work email and create a password. Sigma does not offer a self-serve permanent free tier; the trial lasts 14 days with full feature access.

After signup, Sigma prompts you to set up your organization. Enter your company name and select your region. Sigma hosts its application layer on AWS; your data never leaves your warehouse during analysis because Sigma sends SQL queries directly to your warehouse and renders the results in the browser.

Step 2: Connect Your Warehouse

From the left sidebar, click Administration, then Connections, then Create Connection.

For Snowflake: Select Snowflake from the connector list. You need four values: Account identifier (found in Snowflake under Admin > Accounts, format is orgname-accountname), warehouse name (the virtual warehouse Sigma will use to run queries), a dedicated Sigma service user, and the user's password or key-pair credentials.

Sigma recommends creating a dedicated service user in Snowflake rather than using a personal account. In Snowflake, run:

CREATE USER sigma_user PASSWORD = 'your_password' DEFAULT_ROLE = sigma_role;
CREATE ROLE sigma_role;
GRANT ROLE sigma_role TO USER sigma_user;
GRANT USAGE ON WAREHOUSE your_warehouse TO ROLE sigma_role;
GRANT USAGE ON DATABASE your_database TO ROLE sigma_role;
GRANT USAGE ON ALL SCHEMAS IN DATABASE your_database TO ROLE sigma_role;
GRANT SELECT ON ALL TABLES IN DATABASE your_database TO ROLE sigma_role;

Back in Sigma, enter the account identifier, warehouse name, sigma_user credentials, and click Save. Sigma runs a test query to verify the connection.

For BigQuery: Select BigQuery, then choose OAuth or service account authentication. For team setups, service account is simpler. Download a JSON key from Google Cloud IAM, upload it in Sigma, and set the project ID. Sigma needs BigQuery Data Viewer and BigQuery Job User roles on the project.

Step 3: Build Your First Workbook

Click New in the top navigation and select Workbook. A blank canvas opens.

Click Add Element, then Table. Sigma asks you to select a data source. Browse to your connected warehouse, expand the schema, and click any table to open it in the workbook. Sigma immediately shows you a live preview of the first rows.

From here, the interface works like a spreadsheet. You can:

• Click any column header to sort, filter, or rename it
• Drag columns to reorder them
• Click the plus icon on any column to add a calculated column with formulas like [Revenue] / [Units]
• Right-click any column and select Group By to create a pivot summary

Sigma uses its own formula language that maps closely to Excel syntax. Common formulas work identically: SUM, IF, COUNTIF, VLOOKUP, DATEADD. If you know Excel, most formulas transfer directly.

Step 4: Add Visualizations

Click Add Element again and select Chart. Sigma links the chart to your table data automatically. Select the x-axis column, y-axis column, and chart type from the right panel. Bar, line, scatter, and combo charts are the most commonly used.

All charts update live when you filter the underlying table. You can click a bar in a chart and Sigma will filter the connected table to show only the rows behind that data point. This drill-through behavior is built in; no configuration is required.

Step 5: Share and Publish

Click the Share button in the top right to invite teammates by email. You can set permissions at the workbook level: view only, can comment, or can explore (which lets them add filters and adjust the view without modifying your base workbook).

To create a shareable link with a filter already applied, use Sigma's URL parameters: append ?filter_column_name=value to the workbook URL. This is useful for embedding filtered views in internal tools or Notion pages.

Sigma also supports scheduled exports. Under the workbook menu, select Schedule Export to set up automated CSV or PDF delivery to email addresses or a Slack channel.

Common Mistakes and How to Avoid Them

The most common mistake for new Sigma users is running expensive queries by accident. Because Sigma queries live warehouse data, every column you add or filter you change triggers a new SQL query. On large tables with millions of rows, these queries can run slowly and consume warehouse credits.

To avoid this, apply row-level filters early. Click the filter icon on a date or status column and narrow the dataset before adding calculations. Sigma also has a Preview mode that returns only the first 1,000 rows for fast exploration before you run the full query.

The second common mistake is using a shared admin warehouse instead of a dedicated analytics warehouse. Sigma's queries compete with data loading jobs and other workloads on a shared warehouse. Create a separate Snowflake warehouse sized XS or S exclusively for Sigma; you can auto-suspend it after 60 seconds of inactivity to minimize costs.

What Sigma Does Not Do

Sigma is a consumption layer, not a transformation layer. It does not clean or reshape raw data. If your source tables have inconsistent formats, null values, or naming issues, you need to fix those upstream in dbt, Coalesce, or your warehouse's own transform pipeline before connecting Sigma.

Sigma also does not support offline access or work with data stored in Excel files or local CSVs directly. All data must be in a supported warehouse. If your team's data is still in spreadsheets, a tool like VSLZ can analyze files directly from an upload without requiring a warehouse connection first.

Summary

Sigma Computing is the right tool when your data is already in Snowflake or BigQuery and your analysts need direct access without SQL. The setup takes under 30 minutes. Connect the warehouse, create a workbook, apply filters early to control query costs, and share with teammates using role-based permissions. The learning curve is minimal for anyone with spreadsheet experience.