### Answer Retell AI's Batch Dialer doesn't automatically distribute calls evenly across days—it executes based on your configuration [(1) ]() [(2) ]() . The distribution depends on how you set up the batch call. When creating a batch call, you can configure a

call_time_window

that specifies allowed calling hours and days [(1) ]() . For example, you could set windows for specific days of the week and time ranges . However, the system queues calls without hitting concurrency limits and processes them as quickly as possible within your configured windows [(2) ]() . If you schedule 15,000 calls for a week with a daily calling window (e.g., 8am-9pm) , the actual distribution per day depends on: - Your configured time windows and days [(1) ]() - System capacity and call duration - CPS (Calls-Per-Second) limits if you've set them [(3) ]() For even distribution across 7 days, you'd get approximately 2,143 calls per day (15,000 ÷ 7) . But Retell doesn't enforce this automatically—you'd need to either: 1. Create separate batch calls for each day with ~2,143 numbers each 2. Configure your

call_time_window

strategically to spread the load [(1) ]() The batch calling feature allows you to dispatch thousands of calls by uploading a contact list, with options to "Send Now" or use "Scheduled Time" [(2) ]() .

​ To learn more, see the sources I considered: [[1] Create Batch Call](https://docs.retellai.com/api-references/create-batch-call) [[2] Streamline Calls with Retell AI’s Batch Dialer | Retell AI](https://www.retellai.com/blog/boost-operation-efficiency-with-retell-ais-batch-calling) [[3] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer The inconsistent call volumes you're experiencing (3k → 800 → 550) could be related to several factors based on documented batch calling behavior [(1) ]() [(2) ]() . Potential causes for declining daily volumes: 1. Reserved Concurrency Settings When creating batch calls, you can set a

reserved_concurrency

parameter that reserves slots for non-batch calls (like inbound calls) [(3) ]() [(4) ]() . If this is configured, your batch calls compete for remaining concurrency, which could explain declining throughput. 2. Call Time Window Configuration Your

call_time_window

settings control when calls can execute [(3) ]() [(1) ]() . The declining pattern suggests calls might be hitting window boundaries earlier each day as the queue processes. Verify your configuration includes consistent daily windows across all days [(1) ]() . 3. System-Level Issues There have been documented incidents where scheduled batch calls experienced service disruptions. In January 2026, scheduled batch calls didn't run from 2:35 PM to 5:15 PM PST, with missed scheduling windows not being executed [(5) ]() . While resolved, this indicates batch processing can be affected by system-wide issues. 4. CPS (Calls Per Second) Limits If you've configured CPS limits per provider, this throttles how many calls initiate per second [(6) ]() [(4) ]()

. Default CPS is typically 1 for most providers, which could bottleneck large batches. Recommended actions: 1. Check your batch call's

reserved_concurrency

setting [(3) ]() 2. Verify your

call_time_window

configuration is consistent across all days [(3) ]() [(1) ]() 3. Review your CPS settings if you've configured custom telephony [(6) ]() 4. Monitor the batch call status in your dashboard to see if calls are being queued vs. executed [(1) ]() The batch calling system should process calls as concurrency becomes available within your configured windows [(2) ]() , so significant day-to-day variance without configuration changes warrants investigation.

​ To learn more, see the sources I considered: [[1] Create Batch calls](https://docs.retellai.com/deploy/make-batch-call) [[2] Streamline Calls with Retell AI’s Batch Dialer | Retell AI](https://www.retellai.com/blog/boost-operation-efficiency-with-retell-ais-batch-calling) [[3] Create Batch Call](https://docs.retellai.com/api-references/create-batch-call) [[4] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) [[5] Scheduled batch calls not running](https://community.retellai.com/t/scheduled-batch-calls-not-running/731) [[6] Outbound Calls (Make Calls)](https://docs.retellai.com/deploy/outbound-call) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

### Answer Based on your configuration, here's what's likely causing the declining call volumes: Your Available Batch Concurrency With 225 total concurrency and 150 reserved concurrency, your batch calls have access to only 75 concurrent slots (225 - 150 = 75) [(1) ]() . This is a significant bottleneck for processing 15,000 calls. Why Volumes Are Declining Batch calls are processed as concurrency becomes available within your configured time windows . With only 75 slots available for batch calling: - **Day 1**: Fresh queue, maximum throughput (~3k calls) - **Day 2**: Reduced throughput (800 calls) - likely hitting CPS limits or call duration bottlenecks - **Day 3**: Further decline (550 calls) - queue processing slowing as remaining numbers may have longer connection times or more failed attempts Time Window Configuration Issue Your 8am-8pm window across all days is correct in principle , but with limited concurrency, you're not maximizing the 12-hour daily window effectively. Recommendations 1. **Increase Available Concurrency**: Consider reducing reserved concurrency or upgrading your total concurrency limit [(2) ]() . With 75 slots and average 3-5 minute calls, you can theoretically complete 900-1,500 calls per 12-hour window at full utilization. 2. **Check CPS Limits**: Default CPS is 1 call per second [(3) ]() [(4) ]() . For custom telephony, maximum CPS equals your concurrency ÷ 20 [(5) ]() . At 225 concurrency, your max CPS would be ~11, but this only applies if configured. 3. **Monitor Batch Status**: Check your dashboard to see if calls are queued vs. actively executing [(1) ](

etellai.com/deploy/make-batch-call) . The declining pattern suggests queue processing issues rather than configuration problems. Expected Throughput Calculation With 75 concurrent slots and 5-minute average calls: - 75 calls every 5 minutes = 900 calls per hour - Over 12 hours (8am-8pm) = ~10,800 theoretical maximum daily Your actual Day 1 performance (3k) suggests you're achieving roughly 33% of theoretical maximum, likely due to connection failures, no-answers, or CPS throttling.

​ To learn more, see the sources I considered: [[1] Create Batch calls](https://docs.retellai.com/deploy/make-batch-call) [[2] Understand concurrency & limits](https://docs.retellai.com/deploy/concurrency) [[3] Outbound Calls (Make Calls)](https://docs.retellai.com/deploy/outbound-call) [[4] Platform Changelogs | Retell AI](https://www.retellai.com/changelog) [[5] concurrency calls](https://discord.com/channels/1186424376300679209/1460538188576592017) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.

@AI Evy @can I modify the batch call now to reserved concurrency of 25

### Answer Based on the available documentation, I don't have specific information about whether you can modify the

reserved_concurrency

parameter of an already-created batch call [(1) ]() [(2) ]() . The Create Batch Call API documentation shows

reserved_concurrency

as a parameter when creating a new batch call [(1) ]() , but there's no mention of an update or modify endpoint for existing batch calls in the sources provided. Your options: 1. Check if there's an Update Batch Call API - The documentation I have access to doesn't include an update endpoint, but one may exist 2. Cancel and recreate - You could potentially cancel the current batch call and create a new one with

reserved_concurrency: 25

3. Let it complete - Allow the current batch to finish with your existing settings Important consideration: Reducing reserved concurrency from 150 to 25 would give your batch calls 200 concurrent slots (225 - 25 = 200) instead of the current 75, which should significantly improve your daily throughput [(3) ]() [(2) ]() . I'd recommend reaching out to Retell support directly to confirm whether batch calls can be modified after creation, as this specific capability isn't documented in the sources I have access to.

​ To learn more, see the sources I considered: [[1] Create Batch Call](https://docs.retellai.com/api-references/create-batch-call) [[2] Create Batch calls](https://docs.retellai.com/deploy/make-batch-call) [[3] Understand concurrency & limits](https://docs.retellai.com/deploy/concurrency) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.