Monitoring Benefit
The Monthly view of the Simplified Teams Call Quality Dashboard reports the same reliability and quality metrics as the Hourly and Daily views, aggregated by month.
Its value is different from the other two. Hourly and Daily are for answering "is something wrong?" — Monthly is for answering "is this getting worse?" A tenant can pass every daily threshold for a year while quality declines steadily the whole time, and only the monthly view makes that visible.
That makes this the view to use for capacity planning, for justifying a circuit upgrade, and for showing whether a change actually improved anything.
The three periods answer different questions:
- Hourly — is something wrong right now?
- Daily — is today worse than normal?
- Monthly — this page. Is call quality getting better or worse over time?
The page can be opened directly on the ENow web server:
http://localhost:20080/MailscapeWeb/MicrosoftTeamsCQDMonthly.aspx?server=Office-365
Replace localhost with the web server's host name if browsing from elsewhere, and 20080 with your own port if it was changed at installation.
What is measured
The metric set is the same across all three periods and is documented in full on the Daily page:
- Reliability — Total Call Setup Failure Percentage, Setup Failure Rate Percentage, Total Drop Call Percentage
- Quality — Audio and Video Poor Stream Counts, Average Round Trip, Average Packet Loss Rate, Average Jitter
- Rate My Call — users' own ratings, including a Tier 1 executive list
Reading a monthly trend
A few things are worth knowing before drawing conclusions from a month-on-month comparison.
Call volume changes the picture. Poor stream counts rise naturally as an organisation makes more calls. A month with 20% more calls and 20% more poor streams has not got worse. Read the counts against volume, and lean on the rate and percentage metrics for like-for-like comparison.
Short months and holiday months distort. December and August typically carry lower volume and a different mix of calls — more people at home, more mobile. Compare against the same month last year as well as last month.
The current month is always incomplete. Until it closes it represents a partial sample, and comparing a partial month against complete ones will mislead.
A step change matters more than drift. Gradual movement across several months is usually growth catching up with capacity. A single month that jumps and stays at the new level is a change — a migration, a new site, a network reconfiguration. Find the month, then use the Daily view inside it to find the day.
A gap in the middle of a trend is a collection gap, not a quiet month. If the page reports that no test results have been received for a period, no data arrived for it — which is a different thing from a month with low call volume, and it makes the months either side of it non-comparable. Establish which of the two you are looking at before reading a trend across the gap.
Where the thresholds are set
Warning and critical thresholds are configured in the Office 365 thresholds section of the ENow Admin Console, and are per-install.
Be deliberate about thresholds on this view. Monthly averages are heavily smoothed, so a threshold tuned for daily variation will rarely trigger here — and if it does, the underlying problem is usually severe and has been running for weeks. Many organisations treat the monthly view as a reporting and planning tool rather than an alerting one.
How do we verify the monitoring test results?
- Compare against Microsoft's Call Quality Dashboard at cqd.teams.microsoft.com over the same months.
- Check the reporting window. Microsoft retains call quality data for a limited period, so months beyond that window will not be available to compare against, in either tool.
- Confirm the month is complete before treating it as a data point in a trend.
- Separate "no results received" from "no calls made". The first means collection did not run and nothing was evaluated; the second is a real, quiet month. They look alike here and lead to opposite conclusions.
Common warning or error results, and potential solutions
| Result | Potential solution |
|---|---|
| Poor stream counts rising month on month | Check call volume first. If volume is flat and counts are rising, that is genuine degradation and worth investigating with the Daily view. |
| A single month far worse than its neighbours | Drill into that month with the Daily view. A single bad week, or a single bad site, usually explains it. |
| The current month looks anomalous | It is incomplete. Wait for it to close before comparing. |
| Metrics improved after a change, then drifted back | Common where a capacity increase is absorbed by growth. The monthly view is the right evidence for that conversation. |
| Monthly is green while users complain | Expected — monthly averaging hides short and localised problems. Use Hourly and Daily for live issues, and the VIP Monitoring view for named users. |
| The page reports that no test results have been received | Not a call quality fault. Collection has not run, so nothing has been evaluated — see the data collection checks in Microsoft Teams - VIP Monitoring Status. The Teams views share a collection pipeline and normally fail together. |
Comments
0 comments
Article is closed for comments.