Quick Reference Table
Feature | Opportunity Objects | Account Objects |
Max records per view | 2,000 | 2,000 |
Run for Current Page (max) | 100 records | 100 records |
Run for All (max) | 1,000 records | ❌ Not supported |
Bulk execution via support | ✅ Available | ✅ Available |
Execution mode | Sequential (queued) | Sequential (queued) |
View Record Limits
Maximum Records Per View: 2,000
Each view can display and process up to 2,000 records maximum, even if the total record count shows a higher number.
What this means:
Views are limited to the first 20 pages (with 100 records per page)
If your view contains more than 2,000 records, you won't be able to access records beyond page 20
The total count displayed may show more than 2,000, but only the first 2,000 are accessible
Best practice: Break large datasets into multiple views using filters to keep each view under 2,000 records.
Smart Column Execution Limit
Run for Current Page
Maximum records: 100 records per page
Supported for: Both Opportunity and Account objects
Use case: Quick analysis of records currently visible on your page
Run for All
Opportunity Objects
Maximum records: 1,000 records per execution
Availability: ✅ Supported
Recommendation: For datasets larger than 1,000 records, create multiple filtered views and run each separately
Account Objects
Availability: ❌ Not currently supported
Why: Account views involve complex data mapping (child accounts and related opportunities), which creates larger payloads that can impact UI performance
Workaround: Use "Run for Current Page" or contact Rattle support for bulk execution assistance.
Execution Queue & Rate Limiting
How Smart Column Execution Works
When you initiate multiple runs across different views, Rattle processes them sequentially, not in parallel.
Execution order:
First column completes for all records
Second column starts and completes
Process continues column-by-column
What you'll see:
A "Queued" status for columns and records waiting to start
Columns execute one at a time due to backend rate limiting and batching
Why: Rate limiting ensures system stability and prevents performance degradation for all users.
Long-Running Executions
40-Minute Timeout (UI Only)
If a run takes longer than 40 minutes to complete:
The UI will stop displaying execution progress
You'll see a message indicating the run is still processing in the background
The execution continues on the backend until completion
What to do:
Note the message confirming backend processing
Refresh the page after some time to view the latest generated data
Results will appear once processing completes
Tip: This is not an error—your data is being processed. The timeout only affects UI display, not the actual execution.
Account View Complexity
Why Account View Execution Takes Longer
Account object views are more complex than Opportunity views due to:
Child account relationships that need to be mapped
Associated opportunities linked to each account
Larger dataset payloads that impact UI rendering
Result: Smart column execution on Account views may take longer and consume more system resources.
Best practice:
Use filters to reduce dataset size
Contact support for large bulk executions
Historical Data Backfill
Call & Email Ingestions
Rattle can backfill your historical calls and emails to enable analysis of past interactions.
Important considerations:
Timeline depends on: Number of calls and emails to be ingested
Performance impact: This is a resource-intensive process
Execution window: Performed during off-peak hours only to minimize customer impact
Planning required: Discuss timeline with Rattle team before starting
How to request:
Contact Rattle Support or your Customer Success Manager (CSM)
Provide the approximate number of calls/emails to backfill
Agree on a timeline and execution window
Rattle team will schedule and execute during off-hours
Bulk Smart Column Execution
For Large Datasets (Over Limits)
If you need to run smart columns on datasets exceeding standard limits:
Opportunity: More than 1,000 records
Account: More than 100 records
Process:
Create multiple views with filters to ensure each view contains ≤2,000 records
Contact Rattle Support or your CSM to request bulk execution
Rattle team will run the process as a backend task
Important notes:
Each view is still limited to 2,000 records maximum
Timeline depends on:
Number of records to execute
Amount of data (calls/emails) to ingest for analysis
Execution window: Off-peak hours only due to performance impact
Planning required: Discuss timeline with Rattle team before starting
Example setup: If you have 8,000 opportunity records:
Create 4 views with 2,000 records each using filters
Request bulk execution from Rattle team
Each view will be processed sequentially during off-hours
Credits & Usage
How Credits Are Consumed
Rattle Wizard uses credits for two primary activities:
1 Credit per Call Transcript Analysis
Each call transcript analyzed to generate insights consumes 1 credit
Applies to smart columns analyzing call data
1 Credit per External Research API Call
Online research queries (web searches, data enrichment) consume 1 credit per API call
Applies to smart columns performing external research
Monitoring usage: Check your credit balance and usage history in your board credits used section to track consumption.
Best Practices Summary
✅ Keep views under 2,000 records by using filters.
✅ Break large datasets into multiple filtered views for easier management.
✅ Plan ahead for backfills and bulk runs by discussing timelines with Rattle team.
✅ Monitor credit usage regularly to avoid unexpected depletion.
✅ Use "Run for Current Page" for quick ad-hoc analysis.