ACS error 5000: Message failed to deliver, Please reach out Microsoft support team for more details

Question

ACS error 5000: Message failed to deliver, Please reach out Microsoft support team for more details

Justin Thomas 5

We're getting the above message for many of our ACS SMS messages. Its only started happening today when we attempted to send over 4000 SMS. In particular my SMS number is not working even through my collogues have sent an SMS to that number from their phones and it worked.

Anonymous

2025-11-11T14:48:00.5033333+00:00

Hi Justin Thomas

It may be due to throttling limits when sending a large number of SMS at once.

You can try sending the messages in smaller batches, like 25–25 at a time. That usually helps avoid delivery failures.

Moreover, our internal team will review this further based on the details you’ve shared.
Anonymous

2025-12-02T01:48:52.5366667+00:00

@Justin Thomas

My apologies for the inconvenience. As per our lead’s guidance, a separate azure support ticket from portal is required for this issue.
Justin Thomas 5 Reputation points

2025-12-03T07:40:16.7933333+00:00

As I asked in the private emails - WHERE do I raise this issue?

You have so many different places to do this. "azure support ticket from portal" is not very helpful as this was the place I was told to raise the issue with yourselves after attempting to raise it in another area of Azure.

I also really dislike the 'sneaky' way you all update these.

We were doing all correspondence via email and suddenly, when you can no longer help you come back here and update. So I have to keep an eye on 2 places to see if you've updated me or not.
Anonymous

2025-12-08T20:16:13.6333333+00:00

Hi @Justin Thomas

Internal ACS team made some changes on 4th December. Could you please check and let us know if you are still experiencing the issue?

Thank you for your patience.
Anonymous

2025-12-10T00:11:17.6933333+00:00

Hi @Justin Thomas

Following up to check, is your issue resolved?
Justin Thomas 5 Reputation points

2025-12-10T06:07:46.3333333+00:00

No, we are still getting that failure.

We reduced the volume of messages going out via your service as it was so unreliable. Yesterday (9th Dec) we sent out 48 SMS and had a single failure of the "Message Failed to Deliver" variety.

I cannot say if the problem is fixed. I would not expect any failures. I can attach the send an queue log of the failed SMS to the private messages if that would help?
Justin Thomas 5 Reputation points

2025-12-10T09:28:32.1466667+00:00

No, its still happening today too (10th Dec).

Out of 81 SMS sent 5 gave the '5000: Message failed to deliver, Please reach out Microsoft support team for more details' error.
Anonymous

2025-12-10T18:19:52.61+00:00

Hi @Justin Thomas

thank you for the confirmation, I'll escalate it to internal ACS team.
Justin Thomas 5 Reputation points

2025-12-11T09:23:24.5733333+00:00

And today, we sent 133 of which 17 failed with the "5000: Message failed to deliver, Please reach out Microsoft support team for more details" error.
Justin Thomas 5 Reputation points

2025-12-12T11:35:30.1766667+00:00

Today, we sent 76 and 4 failed with the "5000: Message failed to deliver, Please reach out Microsoft support team for more details" error.
Shree Hima Bindu Maganti 7,590 Reputation points Microsoft External Staff Moderator

2025-12-16T09:37:59.9633333+00:00

Hi @Justin Thomas
The ACS error 5000 (“Message failed to deliver”) in this case is caused by SMS throttling and carrier-side delivery limits when sending messages in high volume or bursts through Azure Communication Services. Even after reducing volume, intermittent failures can still occur because ACS applies dynamic throttling and downstream carrier filtering, which may reject some messages without a more specific error code. The recommended mitigation is to send SMS in smaller batches (for example, 20–25 messages at a time), add delays between batches, and avoid sudden spikes, which significantly reduces failures. Since this issue persisted even at low volumes, it was escalated internally, indicating a service-side or carrier routing issue, not an application bug.
https://learn.microsoft.com/en-us/azure/communication-services/concepts/service-limits
Let me know if you have any further assistances needed.
Justin Thomas 5 Reputation points

2025-12-16T11:17:47.0866667+00:00

For anyone else evaluating ACS SMS:

In our production usage we observed that ACS will accept SMS requests at the API level but may subsequently fail delivery with error code 5000 when internal throughput or burst limits are exceeded. This occurs without HTTP 429 responses or Retry-After guidance.

We only became aware of this behaviour by monitoring delivery queues (or otherwise inspecting post-send delivery status), as the initial API call itself may succeed.

We also observed that once a recipient number enters a 5000 failure state, it can take a significant amount of time (potentially days) before delivery attempts succeed again.

As a result, clients must implement their own batching, throttling, and failure-suppression logic to avoid delivery failures or recipient blocking. This behaviour differs from other SMS providers we have previously used, which enforce rate limits explicitly at the API boundary.

We are continuing to test mitigations as suggested above, but wanted to highlight this behaviour so others can factor it into their design decisions.

Your answer

Anonymous

2025-11-11T14:48:00.5033333+00:00

Hi Justin Thomas

It may be due to throttling limits when sending a large number of SMS at once.

You can try sending the messages in smaller batches, like 25–25 at a time. That usually helps avoid delivery failures.

Moreover, our internal team will review this further based on the details you’ve shared.
Anonymous

2025-12-02T01:48:52.5366667+00:00

@Justin Thomas

My apologies for the inconvenience. As per our lead’s guidance, a separate azure support ticket from portal is required for this issue.
Justin Thomas 5 Reputation points

2025-12-03T07:40:16.7933333+00:00

As I asked in the private emails - WHERE do I raise this issue?

You have so many different places to do this. "azure support ticket from portal" is not very helpful as this was the place I was told to raise the issue with yourselves after attempting to raise it in another area of Azure.

I also really dislike the 'sneaky' way you all update these.

We were doing all correspondence via email and suddenly, when you can no longer help you come back here and update. So I have to keep an eye on 2 places to see if you've updated me or not.
Anonymous

2025-12-08T20:16:13.6333333+00:00

Hi @Justin Thomas

Internal ACS team made some changes on 4th December. Could you please check and let us know if you are still experiencing the issue?

Thank you for your patience.
Anonymous

2025-12-10T00:11:17.6933333+00:00

Hi @Justin Thomas

Following up to check, is your issue resolved?
Justin Thomas 5 Reputation points

2025-12-10T06:07:46.3333333+00:00

No, we are still getting that failure.

We reduced the volume of messages going out via your service as it was so unreliable. Yesterday (9th Dec) we sent out 48 SMS and had a single failure of the "Message Failed to Deliver" variety.

I cannot say if the problem is fixed. I would not expect any failures. I can attach the send an queue log of the failed SMS to the private messages if that would help?
Justin Thomas 5 Reputation points

2025-12-10T09:28:32.1466667+00:00

No, its still happening today too (10th Dec).

Out of 81 SMS sent 5 gave the '5000: Message failed to deliver, Please reach out Microsoft support team for more details' error.
Anonymous

2025-12-10T18:19:52.61+00:00

Hi @Justin Thomas

thank you for the confirmation, I'll escalate it to internal ACS team.
Justin Thomas 5 Reputation points

2025-12-11T09:23:24.5733333+00:00

And today, we sent 133 of which 17 failed with the "5000: Message failed to deliver, Please reach out Microsoft support team for more details" error.
Justin Thomas 5 Reputation points

2025-12-12T11:35:30.1766667+00:00

Today, we sent 76 and 4 failed with the "5000: Message failed to deliver, Please reach out Microsoft support team for more details" error.
Shree Hima Bindu Maganti 7,590 Reputation points Microsoft External Staff Moderator

2025-12-16T09:37:59.9633333+00:00

Hi @Justin Thomas
The ACS error 5000 (“Message failed to deliver”) in this case is caused by SMS throttling and carrier-side delivery limits when sending messages in high volume or bursts through Azure Communication Services. Even after reducing volume, intermittent failures can still occur because ACS applies dynamic throttling and downstream carrier filtering, which may reject some messages without a more specific error code. The recommended mitigation is to send SMS in smaller batches (for example, 20–25 messages at a time), add delays between batches, and avoid sudden spikes, which significantly reduces failures. Since this issue persisted even at low volumes, it was escalated internally, indicating a service-side or carrier routing issue, not an application bug.
https://learn.microsoft.com/en-us/azure/communication-services/concepts/service-limits
Let me know if you have any further assistances needed.
Justin Thomas 5 Reputation points

2025-12-16T11:17:47.0866667+00:00

For anyone else evaluating ACS SMS:

In our production usage we observed that ACS will accept SMS requests at the API level but may subsequently fail delivery with error code 5000 when internal throughput or burst limits are exceeded. This occurs without HTTP 429 responses or Retry-After guidance.

We only became aware of this behaviour by monitoring delivery queues (or otherwise inspecting post-send delivery status), as the initial API call itself may succeed.

We also observed that once a recipient number enters a 5000 failure state, it can take a significant amount of time (potentially days) before delivery attempts succeed again.

As a result, clients must implement their own batching, throttling, and failure-suppression logic to avoid delivery failures or recipient blocking. This behaviour differs from other SMS providers we have previously used, which enforce rate limits explicitly at the API boundary.

We are continuing to test mitigations as suggested above, but wanted to highlight this behaviour so others can factor it into their design decisions.