Skip to main content

Value Suggestions

Querylab.io suggests actual values from your tables when writing WHERE, HAVING, and IN clauses, based on sampled data from BigQuery.

How Value Suggestions Work​

When you type inside a string literal in a filter condition, the editor fetches sample values from the referenced column:

SELECT * FROM events
WHERE event_type = '|'

As you type inside the quotes, completions show actual values from the event_type column:

  • 'page_view'
  • 'click'
  • 'purchase'
  • 'signup'

Press Tab to accept a value.

Where Value Suggestions Appear​

WHERE Clause Comparisons​

WHERE status = '|'

Shows values: 'active', 'inactive', 'pending', 'suspended'

HAVING Clause​

SELECT country, COUNT(*) as total
FROM users
GROUP BY country
HAVING country = '|'

Shows country values from the users table.

IN Clause​

WHERE status IN ('active', '|')

Shows remaining values (excluding 'active' which is already in the list):

  • 'inactive'
  • 'pending'
  • 'suspended'

Comparison Operators​

Value suggestions work with these operators:

  • = (equals)
  • != or <> (not equals)
  • IN (list membership)
  • NOT IN (exclusion)

String Values Only (Currently)​

Value suggestions currently work for string literals only:

-- Works
WHERE status = '|'

-- Not yet supported
WHERE priority = |  (numeric values)
WHERE created_date = |  (date values)

Numeric and date value suggestions are planned for a future update.

Sample Data Source​

Value suggestions use sample data from BigQuery, not the full dataset:

  • No query cost: Fetches sample data from BigQuery at no query cost
  • Sample size: Up to 100-200 rows
  • Freshness: Sample data updates daily

The samples are representative but may not include all distinct values, especially for:

  • High-cardinality columns (thousands of unique values)
  • Recently added values
  • Rare values

Filtering As You Type​

The completion list filters as you type inside the quotes:

WHERE event_type = 'pa|'

Shows only values starting with "pa":

  • 'page_view'
  • 'page_load'

Case-insensitive matching helps find values quickly.

Excluding Existing Values​

In IN clauses, the editor filters out values already present:

WHERE status IN ('active', 'pending', '|')

Completions show only:

  • 'inactive'
  • 'suspended'

(Not 'active' or 'pending' since they're already in the list)

Performance and Caching​

First Request​

The first time you request values for a column:

  • Takes a moment to load (network fetch from BigQuery)
  • Loading indicator: Completion dropdown shows "Loading..."

Subsequent Requests​

After the initial fetch:

  • Instant (results are saved locally)

Prefetching​

The editor can prefetch values when you complete the FROM clause:

FROM events|   [Table name accepted]

Immediately starts loading sample data for common filter columns, so values are ready when you reach the WHERE clause.

High-Cardinality Columns​

For columns with too many distinct values (>1000), value suggestions may not appear:

WHERE user_id = '|'

No suggestions (user_id has millions of unique values). The editor detects high cardinality and skips value completions to avoid overwhelming you with options.

NULL Values​

If a column contains NULL values, the editor suggests using IS NULL instead of = NULL:

WHERE status = |

Shows:

  • IS NULL (special completion)
  • 'active', 'inactive', etc. (string values)

Selecting IS NULL replaces = with IS:

WHERE status IS NULL

Data Freshness​

Sample values are refreshed automatically when:

  • You run a DDL statement (CREATE, ALTER, DROP) on the table
  • You manually refresh the table schema in the tree view
  • Daily cache expiration occurs

Data Privacy​

Value suggestions use the same authentication and permissions as table preview:

  • Only shows data you have access to
  • Uses your Google Cloud credentials
  • No data leaves your browser except to/from BigQuery APIs

Sample values are stored locally in your browser, not sent to any third-party service.

Limitations​

Current Limitations​

  1. String values only - Numeric and date values not yet supported
  2. Tables in FROM clause only - No cross-table value suggestions
  3. Sample data - May miss rare values
  4. No LIKE patterns - Doesn't suggest wildcard patterns like 'prefix%'

Future Improvements​

These features may be added in future updates:

  • Numeric value suggestions
  • Date value suggestions
  • Smart suggestions for LIKE patterns
  • Frequency-based ranking

Disabling Value Suggestions​

Value suggestions cannot be disabled independently, but they only appear when you type inside string literals in filter conditions. Simply ignore them if you prefer typing values manually.