Please note:
The Network Diagnostic Dashboard (available through the Analytics+ plugin) contains even more details about calls with quality issues. For example, you can click on the poor and good Network calls on the charts to inspect the details related to these calls. Learn more here.
___________________________________________________
The Network Diagnostics Dashboard provides comprehensive insights into the network quality and performance of your calls, enabling you to identify and address issues proactively.
You will be able to identify:
- The overall distribution of call network quality (excellent, good, poor) over a filtered time interval.
- The percentage and specific instances of calls experiencing high packet loss, jitter, latency, and ICE failures.
- Detailed diagnostics by IP and user, showing poor network calls categorised by quality metrics.
With this information, you will be able to make informed improvements. For example:
- High Latency: If a significant percentage of calls show high latency (RTT above 300ms), consider investigating network routing.
- Packet Loss: High packet loss rates might indicate issues with network reliability or congestion. Check your network infrastructure and consider quality of service (QoS) improvements.
- Jitter: Calls with high jitter can lead to poor audio quality. Ensure your network is stable and not experiencing excessive variability in packet arrival times.
- User Performance: If specific users consistently experience poor audio quality, consider reviewing their network setup based on their IP location.
- Infrastructure Enhancements: Regularly high ICE failures might necessitate reviewing and enhancing your network's NAT traversal and firewall configurations.
By leveraging the insights provided by the Network Diagnostics Dashboard, you can optimize call quality, enhance user satisfaction, and reduce the incidence of support tickets related to network performance issues.
Availability of this dashboard
This dashboard is currently available to companies with the Professional plan, visible to users with admin or supervisor roles.
Terms used in this dashboard
The main terms used in this dashboard are described in the article Analytics: Terms in the Network Diagnostics section.
⚠️ Please note: Only completed calls have quality metrics.
General tips
|
Tip |
Description |
| Hide and show elements of a graph | If you would like to remove a specific metric from a graph, click directly on the legend. Click on the legend again to make it reappear. |
| Export | Read how to export the data here. |
| Tooltips | Hover over the ℹ️ icon to access quick guidelines from within the dashboard itself. |
Overview of this Dashboard
On this dashboard, you can quickly assess your network diagnostics with the summary monitors at the top of the page, providing a snapshot of network quality.
⚠️ Please note: This dashboard specifically provides information about connected calls and excludes internal calls between Aircall users.
⚠️ Please note: A quick definition of each KPI or graph is given as a tooltip:
⚠️ Please note: In this dashboard, you can find a classification of call network quality into three levels. Calls with poor network are characterised by significant issues such as high jitter, latency, packet loss, or a low Mean Opinion Score (MOS) falling below quality thresholds. Calls with good network have acceptable quality, with minor issues in jitter, latency, or packet loss, and an MOS near the quality thresholds. In contrast, calls with excellent network have superior quality, minimal issues, and a high MOS well above the quality thresholds, indicating high user satisfaction. Find more info in TERMS and METRICS
Network diagnostics overview
In this section, the summaries and general statistics of network quality are displayed.
Specifically, you can learn:
- How many calls exhibited poor network quality?
- How is the number of calls with network issues evolving over time?
Network diagnostics summary
Here we can see the counts and percentages of calls with poor, good, or excellent network.
Network diagnostics summary over time
In this section, you can assess the overall trends and performance metrics related to network diagnostics.
Network diagnostics overview KPI’s
|
Metric/Chart |
Description |
| % Calls with poor network | Percentage of calls with poor network. This includes calls with jitter, latency, packet loss or MOS above quality thresholds. |
| % Calls with good network | Percentage of calls with good network, but not excellent. This includes calls with jitter, latency, packet loss or MOS close to but not exceeding quality thresholds. |
| % Calls with excellent network | Percentage of calls with excellent network quality. This includes calls with low jitter, latency, packet loss, and a high Mean Opinion Score (MOS) well above the quality thresholds. |
Network issues breakdown
In Network issues breakdown section you can see a detailed analysis of the network factors impacting call quality, include key metrics details jitter, packet loss and latency for poor and good network calls
Specifically, you can learn:
- What are the network KPIs associated with calls that have poor o good network quality?
- What are the current weaknesses of your network, or which aspects are at risk of becoming weaknesses?
Calls with poor network - Breakdown
In this section you can see the detail by network KPIs of calls with poor network.
Calls with good network - Breakdown
In this section you can see the detail by network KPIs of calls with good network.
Network issues breakdown KPI’s
|
Metric/Chart |
Description |
| % Calls with high packet loss | Percentage of calls that experienced a packet loss rate exceeding the acceptable threshold of 1.5% or received a high packet loss flag alert during the call. |
| % Calls with high jitter | Percentage of calls that experienced jitter above the acceptable threshold of 30 milliseconds or received a high jitter flag alert during the call. |
| % Calls with high latency | Percentage of calls that experienced RTT above the acceptable threshold of 300ms or received a high latency flag alert during the call. |
| % Calls with medium packet loss | Percentage of calls that experienced a packet loss rate greater than 1% but less than the 1.5% threshold. |
| % Calls with medium jitter | Percentage of calls that experienced jitter rate greater than 20 ms but less than the 30 ms threshold. |
| %Calls with medium latency | Percentage of calls that experienced RTT rate greater than 200 ms but less than the 300 ms threshold. |
Learn where the issues come from
You can then use this section to delve into specific issues affecting your calls. This section helps you identify the root causes of network quality problems, allowing you to respond with targeted actions to improve the overall user experience
In this section, you can identify the reasons for poor network quality as well as their locations based on IP and affected users.
Specifically, you can learn:
- Are calls with poor network coming from a specific external IP?
- Are calls with poor network coming from a specific user on one or more external IPs?
- What number of calls with good network, but not excellent, do I have on different external IPs?
Network diagnostics per IP
In this section you can see the top 15 IPs with the most calls with poor network.
Network diagnostics Breakdown by IP and user
In this section you can see the detail of network KPIs in calls with poor network by IP and user.
⚠️ Please note: Each IP address corresponds to an exact geographical location for user connections. This location data allows us to identify the specific network environment and physical location of the client's device during a call. By analyzing this information, we can accurately pinpoint where call quality issues originate.
|
Metric/Chart |
Description |
| External Media IP | Public IP address used to route media traffic during the call. While it typically reflects the device's IP, it can vary when using proxies or TURN servers. |
| User | The specific user who connected with the external caller or call recipient on this call (if any). In cases where the call is transferred from one Aircall number to another Aircall number, only the last user is provided here. |
|
% calls with ICE failure |
Percentage of calls with an error or breakdown occurring during the process of establishing connectivity using the Interactive Connectivity Establishment (ICE) framework |
|
MOS |
The MOS (Mean Opinion Score) is an automated score between 1-5 of your experienced audio quality, represented as follows: 1 → Bad 2 → Poor 3 → Fair 4 → Good 5 → Excellent This is calculated based on the values of latency, packet-loss, and jitter for the duration of the call. |
| % calls with MOS score below 4 | Percentage of calls with MOS (Mean Opinion Score) below 4 over the set time period. |
Filters available in this dashboard
|
Filter |
Description |
|
Date |
The date range you wish the data in this dashboard to reflect. (This will include calls that start within this date range in the selected timezone) |
|
Date Breakdown |
For timeline graphs: the breakdown of the timeline e.g. one point per day, week, or month. |
|
Duration |
Specify the length of calls in seconds. This allows you to analyze and categorize call data based on the precise duration of each call |
|
Users |
Select specific users or agents. This allows you to focus on the call data and performance metrics related to particular individuals within the system. |
|
Timezone |
The timezone that your selected dates will reflect |