Benchling Insights is a data analysis and visualization tool native to the Benchling platform. It enables users to create dashboards, run SQL queries, and visualize data to support scientific, operational, and administrative decision-making. Insights is available to tenants on Enterprise or Industry plans.

Overview of Benchling Insights

Insights organizes SQL queries into blocks, and blocks into dashboards, enabling users  to explore, analyze, and present their data.

• SQL and Queries: queries in Insights are written with postgreSQL, to filter, and join metadata across schemas
• Blocks: Each block contains a query and can display the output as a raw table or as a visualization such as a bar chart, line graph, or regression model.
• Insights Dashboards: a Dashboard is a curated collection of blocks that serve as a centralized, interactive space where users can view, filter, and edit data outputs depending on their permissions. In a dashboard, you can create new blocks, duplicate or rename existing ones, and manage overall layout and content—ensuring data is both actionable and accessible across teams.

Generally when working with Insights you will: 

1. Create a Dashboard
2. Create one or more blocks
3. Filter and visualize data
4. Edit visualizations in blocks or Dashboards 

The sections in this article walk you through these actions

Create a Dashboard

Dashboards are containers for blocks. They organize data and visualizations in one place and let you monitor key metrics across projects. To create a new Dashboard:

1. Click the Insights icon in the navigation bar
2. Click the local create icon and select Dashboard from the menu
3. Use the textbox to enter a dashboard name and select the associated project
4. Click Create 

View a Dashboard

Viewing a dashboard allows you to access the collection of blocks, queries, and visualizations it contains. 

1. Click the Insights icon in the navigation bar
2. In the panel, locate the dashboard you want to open
3. Click the Dashboard name to open it

Create a Block

Blocks are the core work units of a Dashboard. They run queries against Benchling’s data warehouse and display results as tables or visualizations.Creating a block lets you add a new query or visualization to a dashboard, turning raw data into actionable insights. Blocks can be created in three ways:

• Block template – Configure lists of entities from specific schemas and fields; best for quick setup without SQL knowledge.
• Custom SQL query – Write queries that join schemas, filter data, and customize how query results are displayed; best when you need precision and complex logic.
• Duplicate an existing block – Reuse and adapt an existing query or visualization; best when you want to build on work that already exists.

Note: If you are unfamiliar with SQL, consider creating a new block from a template instead of editing an existing query.

Block Templates

Block templates provide a guided, no-code way to build queries. You can select data, fields, and filters through the interface to quickly set up standard tables or visualizations without writing SQL. 

1. Navigate to the desired dashboard
2. Click Add block
3. Choose List of entities and follow the numbered sections to build your query and select your data of interest
4. Use the Select data dropdown to select which schema you would like to use in your query
   • If you anticipate there might be duplicate values that you want to remove from the results, use the checkbox to exclude duplicate values
5. In the Choose fields to include section, use the checkboxes to toggle which field(s) you want displayed
   • To change the display name for this query, use the textbox next to the field of interest and update it to the new name
6. In the Filter results section, use the dropdown to add filter(s) to your data. Select the attribute from the menu
   • A new filter dropdown will appear, decide if you want the filter to be set to is one of or is none of
   • Define the filter attribute with using the textbox
7. Click Run Query to save what you’ve built and to display results of the query

Custom SQL

Custom SQL blocks allow you to write and execute SQL queries directly against Benchling’s data warehouse. This allows you to join multiple schemas and build complex, reusable queries. To create a custom SQL block:

1. Navigate to the desired dashboard
2. Click Add block
3. Choose SQL query
4. Enter your SQL code in the editor
5. Click Run Query

Edit a Block

Editing a block allows you to update the SQL query, adjust parameters, or change the visualization type used to display your data. The list below discusses some of the key tools of the query editor: 

• SQL Editor: Write SQL queries, run them against Benchling’s data warehouse, and view results.
• Schema Browser: Explore Benchling’s data structure. Search for tables and fields, copy exact names into queries
• Entity Relationship Diagram (ERD): Shows how tables relate to each other. Understand which fields connect tables and design accurate joins.

To edit a block: 

1. Navigate to the block you want to edit
2. Click the edit icon in the block’s top-right corner
3. The SQL editing window will open
4. Edit the SQL query directly, update parameters, or adjust the visualization type
5. Click Run Query to apply your changes

Search for schemas with the schema browser

The schema browser helps you locate the correct tables and columns to use in your SQL queries. Using it ensures you’re pulling from the right data source and reduces query errors caused by typos or mismatched names.

1. Enter a search term in the schema browser search bar
2. Click a table name to expand it
3. Scroll left to right to view the full list of columns in that table
4. Highlight and copy a table or column name to paste it into your SQL query

Note: Be sure to use the exact column header when referencing it in your query. Even small differences in spelling, capitalization, or underscores will cause the query to fail.

Explore the entity relationship diagram

The Entity Relationship Diagram (ERD) viewer helps you visualize how tables in Benchling’s data warehouse are structured and connected. In the ERD, you can expand tables, view their columns, and see how they relate to other tables. To use the ERD viewer: 

1. In the schema browser, search for and select a schema
2. Click the Open entity diagram icon to launch the ERD view

About the ERD viewer

• View schema context: Get a visual map of how entities are connected
• Explore table relationships: See how tables are connected through key fields
• Expand table details: Click on See more columns at the bottom of a table
• Identify join keys: Use the relationships to determine which columns link tables together
• Edit SQL in place: Use the SQL query pane on the right to write or refine your query

Filter and visualize data

Configure parameters 

Parameters make dashboards interactive by allowing users to filter or adjust queries without editing SQL. This makes dashboards reusable and customizable across different projects or use cases. To configure a parameter:

1. In your SQL query, replace a fixed value with the parameter name enclosed in double curly brackets (for example: {{Assay Type}})
2. Click Run Query
3. In the dashboard header (above the block editor), click the bolded parameter name to configure the parameter (e.g., Assay Type)
4. From the dropdown, select the parameter type (e.g., text, number, date, entity, dropdown). Check Allow multiple if applicable
5. Click Update to save changes

Apply a parameter to filter a query

When multiple blocks in the same dashboard use a parameter with the same name, that parameter acts as a shared filter across all of them. Setting the parameter value once in the dashboard header automatically applies it to every block that references it.

Shared parameters make dashboards interactive and reusable. For example, a user could filter every block in a dashboard by Cell Line or Assay Type with a single dropdown selection, ensuring consistent filtering across tables and charts. To apply a parameter as a filter:

1. In the dashboard header, click the non-bold side of the parameter button (e.g., where it says no value)
2. Enter a new value or select one from the list of options
3. Click Update to apply the filter

When using parameters, remember: 

• Parameters must match the field type they filter (e.g., text vs. number)
• Using parameters instead of hard-coded values makes queries flexible and dashboards adaptable
• The bolded text in the parameter button is for configuring the parameter
• The non-bolded text is for setting or changing its value

Visualize data

Visualizations transform raw query results into graphs and charts, making it easier to identify patterns, trends, and anomalies. The table below shows the types of visualizations you can create in Insights, and when to use them. 

Create a visualization

1. In the block editor, open the Chart type dropdown in the bottom-right corner
2. Select the chart type you want to use
3. Use the Setup, Format, and Analysis sub-tabs (availability depends on chart type) to customize your visualization.
   • Setup defines axes, series, rows, or columns
   • Format controls chart display, such as ranges, labels, and color styling
   • Analysis applies advanced options like regression or error bars (only on certain charts)
4. When finished, click the dashboard title at the top of the block editor to return to the dashboard view

Sub-tab availability depends on chart type. For example, Pie charts only have Setup and Format, while Scatter plots include Setup, Analysis, and Format.

Use these tabs to refine your visualization. Common adjustments include relabeling axes, formatting numeric values, applying aggregation functions, or highlighting thresholds with color.

Note: You cannot export a chart image directly. Only the underlying table data can be downloaded (CSV) or copied.

Data limits in visualizations

Insights visualizations can render a maximum of 1,000 data points per block. If a query returns more than 1,000 points, the visualization may fail to display. To stay within limits: 

• Apply filters to reduce the dataset size
• Use aggregation (e.g., counts, averages) to summarize results
• Add a LIMIT clause in your SQL query to restrict rows returned

Note: Even if a visualization doesn’t render, you can still view and export the full table of results.

Using dynamic pivots

Dynamic pivots reshape query results by turning rows into columns. This makes it easier to create flexible summary tables, such as showing assay results by condition or notebook entries by month. The dynamic_pivot function is available only on non-GXP tenants.

In your SQL query, use the DYNAMIC_PIVOT function

dynamic_pivot(<NUMBER_OF_ROW_FIELDS>, <DATA_QUERY>, <COLUMN_QUERY>)

1. Define the NUMBER_OF_ROW_FIELDS  how many leading columns from your data should remain as rows (must match exactly).
2. Write the DATA_QUERY to return exactly:
   • The row fields to remain as rows
   • One pivot key: the field whose distinct values will become new columns
   • The values that will populate the table body
3. Write the COLUMN_QUERY to return the distinct values for the new column headers. It should be the same type/format as the pivot key.
4. Run the query to generate your pivoted table

Queries inside dynamic_pivot must be written as strings; wrap them with $$ … $$ for simplicity

Note: You can pivot on only one field at a time

Configure regressions and detect outliers

With scatter plot visualizations, you can model linear regressions for continuous variables with a straight-line relationship or logistic regressions (4PL) for data that follows a sigmoidal curve. 

Outlier detection is only available on logistic regressions, you have several options to select from: 

• Detect – marks outlier data points with a red X; calculations still include them
• Detect and exclude – marks outliers with a semi-transparent red X; calculations exclude them
• Use the sensitivity slider to adjust how aggressively outliers are flagged

Configure a linear regression

1. Navigate to the block you want to edit and click the edit icon
2. From the Chart type dropdown, select Scatter plot (table is the default)
3. In the Setup tab, set the x-axis and y-axis
   • The y-axis must be numeric
   • Leave Series as None
4. Open the Analysis tab and select Linear from the Regression line dropdown
   • If needed, choose an Aggregation option to combine y-values that share the same x-value
   • If needed, add Error bars (standard deviation or standard error)
5. View regression parameters and summary statistics in the top-right of the chart
   • Click the information icon to open a parameter table
   • Click the copy icon to copy regression values

Note: Linear regression is calculated using the ordinary least squares model: y = mx + b.

Configure a logistic regression

1. Navigate to the block you want to edit and click the edit icon
2. From the Chart type dropdown, select Scatter plot (table is the default)
3. In the Setup tab, set the x-axis and y-axis
   • The x-axis and y-axis must be numeric
   • Leave Series as None
4. Open the Analysis tab and select 4-parameter logistic from the Regression line downdown. Additional options:
   • Click the settings icon next to the regression dropdown to open Regression settings
     • Under the Parameters tab, you can optionally constrain values for Min, Max, IC50, or Slope
     • Expand Advanced settings to set initial parameter values (defaults are filled automatically)
   • If needed, choose an Aggregation option to combine y-values that share the same x-value
   • If needed, add Error bars (standard deviation or standard error)
   • If needed, use the Outliers dropdown to flag or exclude anomalous points
5. View regression parameters and summary statistics in the top-right corner of the chart
   • Click the information icon to see a parameter table
   • Click the copy icon to copy regression values

Note: Logistic regressions use the 4-parameter logistic model with least squares fitting. For non-linear models, R² does not always measure goodness-of-fit. Benchling also provides a p-value; lower values imply a more statistically significant fit

Embed Dashboards into the Registry

Embedding a dashboard into a Registry schema creates a new Dashboard tab when viewing an entity. This tab appears alongside other entity views such as Results and Metadata, allowing you to see a snapshot of a dashboard directly in context. 

Embedding lets you review analytics side by side with entity data, streamlining decisions and reducing the need to switch between the Registry and Insights. If parameters are bound, they auto-populate with the current entity. If not, the dashboard shows default results and the user can manually update parameters. 

Note: Access to embedded dashboards follows both Registry and Project permissions.

To embed a Dashboard in a schema:

1. Navigate to Registry Settings by clicking your avatar and selecting Feature Settings
2. Click on the entity schema where you want the Dashboard to appear
3. In the Dashboard section, click the + icon
4. Use the dropdown to select the dashboard you want to embed
5. Configure parameters:
   • If a dashboard parameter matches the entity type of the schema, select Set parameter to entity being viewed. This automatically filters the dashboard to the entity currently open (e.g., when viewing Cell Line A, the dashboard displays data for Cell Line A)
   • For other parameters, enter a default value so the dashboard has data to display immediately
6. Click Save

Note: Entities with the embedded dashboard must be refreshed for changes to take effect. If parameters aren’t bound, the tab opens with defaults; users can still set values manually.

Edit blocks and Dashboards 

Duplicate a Dashboard

Duplicating a dashboard lets you reuse existing configurations and queries without starting from scratch.

1. Click the ⋯ icon in the top-right corner of the dashboard
2. Select Duplicate from the menu

Rename a Dashboard

Renaming a Dashboard allows you to ensure Dashboards have descriptive names to make it easier to identify its purpose and keep your workspace organized.

1. Hover over the dashboard name, and click the edit icon
2. Use the textbox to enter the new name
3. Click the checkmark icon to save the name

Modify a Dashboard layout

Editing a Dashboard allows you to adjust its layout so it better fits your needs. You can reorganize the order of blocks by dragging and dropping them into a new arrangement, making it easier to highlight the most relevant data and visualizations. To edit a Dashboard:

1. Open the Dashboard you want to edit
2. Click and drag the six dot icon on a block and drag it into a new order

Delete a Dashboard 

Deleting a Dashboard permanently removes Dashboards that are outdated, incorrect, or no longer needed.

1. Click the ⋯ icon in the top-right corner of the dashboard or block
2. Select Delete from the menu
3. In the confirmation window, click Delete again

Note: Deletion is permanent. Deleted dashboards and blocks cannot be recovered.

Use block actions to update blocks 

There are several actions you can take from the more options menu to edit and update blocks. These include: 

• Download block as CSV: this lets you export the underlying query results as a CSV file for offline analysis
• Copy block to clipboard: this lets you copy table data for use of pasting into slides or other documents
• Add to worklist: this lets you add the contents of a query directly to a Worklist to streamline downstream information handoffs
• Duplicate block: this lets you reuse an existing query and visualization without starting from scratch, and is useful if you want to make a small change to a query while preserving the original
• Delete a block: this allows you to permanently delete the block and results of the query

Depending on what has been queried, you may have other options including to create or execute a Workflow task or create a request. To access these actions: 

1. In the block that you want to view, click the … icon in the block’s top-right to access the menu
2. Click on the option for the action that you want to take
3. Fill out the information in the modal that opens to complete the action

Note: CSV exports include the raw query results, not what is displayed in a visualization. Visualizations cannot be exported directly, but you can capture them as screenshots.

Rename a block

Renaming a block clarifies its purpose and makes Dashboards easier to scan. Renaming a block doesn’t change its SQL or visualization—only the display name.

1. In the block you want to rename, click the edit icon in the top-right corner
2. In the block editor view, look at the header bar above the editor where the block name appears (e.g., Block 1)
3. Hover over the block name and click to reveal an inline edit icon
4. Use the textbox to type the new name, then click the check icon to save
5. Click back to the dashboard—your block now shows the new name

Optimize queries

Optimizing queries keeps dashboards fast and responsive, reduces resource usage, and helps prevent timeouts or slow performance. To optimize your queries, focus on the following:

Improve query performance 

• Specify columns explicitly instead of using SELECT *
• Use LIMIT to restrict the number of rows returned
• Minimize operations that slow performance, such as ORDER BY, DISTINCT, and UNION

Filter early

• Apply filters with WHERE clauses as soon as possible in your query

Simplify queries

• Split complex queries into smaller, manageable parts
• Avoid using FULL OUTER JOIN unless absolutely necessary

Identifying Bottlenecks

• Use EXPLAIN to view the execution plan and understand how your query runs
• Use EXPLAIN ANALYZE to execute the query and view actual runtime statistics, helping you pinpoint slow steps

Note: If dashboards are running slowly, start by reducing columns, rows, or joins before troubleshooting with EXPLAIN.

For examples of how to optimize a query

Keyboard shortcuts for Insights

Keyboard shortcuts improve productivity in Insights by letting you navigate, edit, and run queries without relying on the mouse. These are especially useful for power users who frequently build or update dashboards. 

Global shortcuts

• Cmd + K / Ctrl + K: Open the command palette
• /: Open search bar from anywhere
• Cmd + Shift + A / Ctrl + Shift + A: Open activity panel
• Cmd + Shift + P / Ctrl + Shift + P: Open pinned items
• Cmd + . / Ctrl + .: Expand/collapse sidebar

Insights-specific shortcuts

• Cmd + Enter / Ctrl + Enter: Run SQL query in block editor
• Esc: Exit full-screen block mode or close pop-ups
• Tab: Navigate between editable elements (e.g., dashboard name, block settings)
• Cmd + C / Ctrl + C: Copy selected text or block output
• Cmd + V / Ctrl + V: Paste text or data

Navigation and editing tips

• Use Tab and Shift + Tab to move through block fields and input areas
• Arrow keys: Navigate dashboards and select elements
• Cmd + Z / Ctrl + Z: Undo changes (where supported)
• Cmd + Shift + Z / Ctrl + Shift + Z: Redo changes

Export an audit log 

Audit logs help tenant admins track how dashboards and blocks are created, modified, and deleted. They provide visibility into user actions for compliance, governance, and troubleshooting. The exported file includes timestamps, users, action types (create, edit, delete), and object identifiers. Audit logs capture: 

Dashboard-Level Events

• Dashboard created
• Blocks added or removed
• Dashboard embedded into a Registry schema

Block-Level Events

• SQL query edits (with old vs. new SQL text)
• Visualization type changes (e.g., Bar Chart → Scatter Plot)
• Visualization setting changes (axes, error bars, formatting, etc.)

Parameter Events

• Parameters created or deleted
• Parameter type updated (e.g., free text → dropdown)
• Parameter values added or removed
• Parameter set to single- or multi-value

To export an audio log:

1. Navigate to the Dashboard you want to audit
2. Click the ⋯ icon in the top-right corner of the dashboard
3. Select Export audit log
4. In the pop-up window, choose an export format (PDF or CSV)
5. Click Export

Note: If the export takes longer than 5 minutes, the file will be emailed to you, otherwise, it will download directly to your computer

Permissions and access 

Permissions in Insights govern who can view, create, and edit dashboards and blocks. Understanding these roles ensures the right people have access to critical information while maintaining data governance.

Benchling’s permissions model is tied to Projects, the Registry, and optional schema-level access policies. This means that a user’s access may differ between projects or embedded dashboards depending on their role. The table below outlines the actions that users with different permissions can take:

How Permissions Work

• Project-scoped: Every Insights dashboard must be linked to a project. A user’s ability to view, edit, or administer the dashboard is inherited from their role in that project
• Registry-embedded dashboards: When dashboards are embedded in Registry schemas, users need access to both the schema and the dashboard’s project. If either is restricted, the dashboard will not be visible
• Schema-level access: Tenant admins can configure schema access policies to further restrict visibility of certain datasets, even if a user has project-level rights
• Global governance: Feature flags and tenant-wide policies may further control whether Insights is available and which datasets can be queried

Generally, we recommend that you align Dashboard access with existing project roles to keep permissions consistent. 

Frequently asked questions

Q: Who can access Benchling Insights?

A: Insights is available to tenants on Enterprise or Industry plans

Q: Can I edit data directly in the data warehouse through Insights?

A: No, Insights does not support editing data in the warehouse

Q: Can I import and analyze external data in Insights?

A: No, Insights does not support importing and analyzing external data

Q: How can I learn more about writing SQL queries for Insights?

A: Benchling offers a SQL basics video series that covers foundational concepts, clauses, and functions commonly used in Insights

Q: How can I improve the performance of my queries?

• Use filters to reduce dataset size
• Avoid unnecessary joins or subqueries
• Use indexed fields where possible
• Aggregate results to minimize return size

Q: Why am I seeing duplicate rows in my query results?

• Joins without unique keys may result in repeated rows
• You may need to use SELECT DISTINCT
• Check for redundancy in underlying tables

Q: Can I use parameters in my queries?

Yes. Parameters allow dynamic query inputs. They can:

• Filter data interactively
• Customize dashboards for different users
• Accept multiple values if configured using IN UNNEST(split({{parameter}}, ','))

Q: How do I handle parameters with multiple values?

Use a comma-separated input and split it:

WHERE my_field IN UNNEST(split({{my_parameter}}, ','))

This allows the user to enter "A,B,C" and match all three values

Q: Can I join data across schemas?

Yes, you can join across tables as long as you use compatible keys. Use JOIN clauses with clear aliases and filtering to avoid duplication

Q: How do I share my dashboards with others?

• Ensure others have access to the associated project
• Use project sharing settings to manage visibility
• Embed dashboards in entity schemas for broader access
• Export charts or tables as CSV for external sharing

Q: What should I do if my query returns no results?

• Confirm that filters are not too narrow
• Check that data exists in the selected tables
• Inspect JOIN conditions for exclusions

Q: Is there a limit to the amount of data I can query?

Yes. Benchling may impose limits for performance reasons. Use filters and aggregations to keep result sets manageable

Q: How can I format numbers in my query output (e.g., rounding, units)?

Use SQL functions such as ROUND(value, decimals) to round numbers. You can also use TO_CHAR() to format values with units, commas, or currency symbols.

Q: What should I do when I encounter SQL errors or unexpected query behavior?

Check for:

• Syntax issues or misspelled table/column names
• Incompatible data types in joins or filters
• Extra or missing commas, quotes, or parentheses Use LIMIT 10 during testing to validate queries on a small result set.

Q: How do I structure complex joins or multi-table queries?

Use clear aliases and always specify join keys explicitly. Start by testing simpler two-table joins, then add more layers. Confirm each intermediate result with LIMIT.

Q: When should I use dynamic_pivot() and what are its limitations?

Use dynamic_pivot() when your data has a variable set of categories you want to convert into columns. Avoid using it with large datasets or in GxP environments, as it's only supported on non-GXP tenants.

Q: How can I debug or analyze query performance?

Use EXPLAIN to view the query plan and EXPLAIN ANALYZE for execution statistics. This helps identify bottlenecks, such as expensive joins or scans.

Q: What are best practices for writing reusable queries?

• Use descriptive aliases and parameter names
• Avoid hardcoded values—use parameters instead
• Structure queries with readable indentation and comments
• Test and version SQL queries using dashboard duplication

Q: What is the difference between a ‘regular’ and a ‘$raw’ table?

A: regular (non-raw) table is a validated set of entities with fewer columns and fewer rows; archived and unreviewed values are excluded. For example, data from Results tables in unreviewed entries could be accessed by querying the $raw tables but not by querying the regular tables.

Q: Why are ids listed in the form “ent_fhgTCjd” and how do I view the name or entity chip?

This is the API ID. You can view the entity chip in Insights by clicking the 'Show Raw Output'/'Hide Raw Output' (table icon) button at the top right of the Insights Output Block, right above the 'Chart Type' dropdown.

Most warehouse tables have a 'name' and a 'registry_id' column. Include this column in your query to view the name or registry.

Q: Why isn’t my chart rendering?

A: Insights visualizations are limited to ~1,000 data points. Apply filters, aggregations, or LIMIT in your query to reduce the dataset.

Q: What happens if I don’t set a parameter value?

A: Dashboards will either show the default value (if configured) or display “No value” until you select one.

Q: Can I run Insights queries through the API?
A: Not directly. While you can’t execute Insights queries via API, you can replicate the logic using Registry or Notebook APIs and external tools.

Q: Why does my query time out?
A: Large or complex queries can exceed execution limits. Try selecting fewer columns, applying filters earlier, or avoiding heavy operations like ORDER BY or DISTINCT.