> ## Documentation Index
> Fetch the complete documentation index at: https://help.berocker.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Messaging health

> What your automated messaging is doing right now — the queue, how each sending number is performing, and the limits behind both.

**Settings → Templates → Health** answers the question every other messaging screen dodges:
*what is actually happening right now?*

> What the automated messaging is doing right now — what is queued and roughly how long it will take
> to clear, how each sending number is really performing, and the limits behind both. Delivery
> figures count only messages the network gave a definite answer about, and how much of the traffic
> that was is shown beside them.

<Frame>
  <img src="https://mintcdn.com/berocker/JlrUOS-Vn_pO8ZdT/images/settings/templates-health.png?fit=max&auto=format&n=JlrUOS-Vn_pO8ZdT&q=85&s=1e6ddb1076ff5c79890828d291bf727b" alt="The Messaging health tab showing the queue tiles, sending numbers table and sending limits" width="3200" height="4800" data-path="images/settings/templates-health.png" />
</Frame>

***

## The four tiles

| Tile                     | What it tells you                                                                     |
| ------------------------ | ------------------------------------------------------------------------------------- |
| **Waiting to send**      | How many messages are queued, with how many are **due now**.                          |
| **Backlog**              | Roughly how long the queue takes to clear, at the current per-number-per-minute rate. |
| **Held for the morning** | Messages the quiet window is holding. They go out when quiet hours open.              |
| **Opted out**            | People asking to be left alone.                                                       |

<Tip>
  A big **held for the morning** number is normal, not a fault — it is tonight's follow-ups queued
  up for the morning release.
</Tip>

***

## Sending numbers, last 24 hours

One row per number, with a **"N need attention"** counter in the corner.

| Column           | What it is                                                      |
| ---------------- | --------------------------------------------------------------- |
| **Number**       | The sending number.                                             |
| **Attempted**    | Messages sent from it.                                          |
| **Delivered**    | Confirmed delivered.                                            |
| **No word yet**  | Sent, no outcome reported.                                      |
| **Failed**       | Confirmed failures.                                             |
| **Failure rate** | Failures as a share of messages with a definite answer.         |
| **Confirmed**    | How much of this number's traffic got a definite answer at all. |

A number in trouble is flagged **Needs attention** in red.

<Info>
  Read the two together, as the screen itself says:

  > The failure rate counts only messages the network gave a definite answer about. "Confirmed" is
  > how much of the traffic that was — a high failure rate over a low confirmed share is a much
  > smaller statement than the same rate over a high one.
</Info>

<Warning>
  A number marked **Needs attention** is on its way to being blocked by carriers. Rest it, check
  what you have been sending from it, and make sure your opt-out line is on.
</Warning>

***

## Sending limits

> Leave a field blank to use the default shown beside it.

| Setting                                       | What it does                                                                                                                                                                                                                            |
| --------------------------------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **Texts per number per minute**               | *Default 4.* Anything over this waits a few minutes and goes out in a later one — never dropped, never counted as a failure. Zero removes the limit.                                                                                    |
| **Quiet hours**                               | Follow-ups and reminders send only between the hours below, in the **customer's** local time. Outside them they hold and go out the next morning.                                                                                       |
| **Sending starts at** / **Last sends by**     | The open window. *Customer's local time, not yours. Default 8 AM to 10 PM.*                                                                                                                                                             |
| **Spread the morning backlog over (minutes)** | *Default 45.* Not a delay — messages held overnight go out between 8:00 and 8:45 instead of all at once on the hour. One number firing its whole backlog in a minute is what gets numbers flagged. Zero releases everything at opening. |
| **Failure rate worth flagging (%)**           | *Default 10%.* A number is only judged once **50** of its messages have a definite answer.                                                                                                                                              |
| **Opt-out line**                              | Added to the first automated text of a conversation and again once a month — not to every message, which would push many of them over a length boundary for nothing.                                                                    |
| **News goes out at night**                    | Quotes, confirmations, dispatches and delivery updates send when they happen. Follow-ups and reminders still wait for the morning.                                                                                                      |
| **Rest a failing number automatically**       | *Off by default.* With it on, a number over the threshold stops sending until it recovers — its messages wait rather than fail. Leave it off to be told first.                                                                          |

Click **Save** when you are done.

<Warning>
  Quiet hours are worked out in the **customer's** local time, using your organization's time zone
  as the base. If the time zone in [Business settings](/settings/business) is wrong, this is wrong
  too.
</Warning>

***

## "The texts are not going out"

Work down this list:

<Steps>
  <Step title="Is it queued, or is it stuck?">
    Check **Waiting to send** and **Backlog**. If the backlog is shrinking, it is going out — just
    not yet.
  </Step>

  <Step title="Is the quiet window holding it?">
    A large **Held for the morning** count means exactly that.
  </Step>

  <Step title="Is a number resting?">
    Look for **Needs attention**, and whether **Rest a failing number automatically** is on.
  </Step>

  <Step title="Is it the pace?">
    A low **texts per number per minute** with a big queue is slow by design. **Backlog** tells you
    how slow.
  </Step>

  <Step title="Is the template even active?">
    Back to [Templates](/settings/templates) — an inactive template never sends.
  </Step>

  <Step title="Does the sender exist?">
    No default SMS contact means nothing sends at all. See [Profile](/settings/profile).
  </Step>
</Steps>

***

## Related

<CardGroup cols={2}>
  <Card title="Templates" icon="message" href="/settings/templates">
    The messages themselves.
  </Card>

  <Card title="Email deliverability" icon="envelope" href="/integrations/email/sendgrid/email-deliveribility">
    The same problem, for email.
  </Card>
</CardGroup>
