This browser is no longer supported.
Upgrade to Microsoft Edge to take advantage of the latest features, security updates, and technical support.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(182.40000000000003, 99%, 27%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3ES%3C/text%3E%3C/svg%3E)
We routinely pull several aggregate (the 30 days and daily) usage metrics from the graph API (e.g., Microsoft Teams user activity, OneDrive activity, SharePoint activity, etc), but I would like understand exactly what time frame (i.e., down to the second) "30 days" includes. Here is a detailed scenario:
Let's say our job kicks off at 9:00 a.m. UTC on 10/1/2024 and we are pulling a 30 day usage report. Does that report account for activity from essentially 9/1/2024 09:00:00 UTC - 10/1/2024 09:00:00 UTC (aka, 30 rolling days from the exact second), or 9/1/2024 00:00:00 UTC - 9/30/2024 23:59:59 UTC (aka, most recent full 30 calendar days)?
No. Reports are generated with few days delay (at best). You can check the reportRefreshDate property within the output to get an idea as to what date the data is current for, as well as the reportDate property for individual datapoints.
Hi Vasil. We're getting report data the next day so I haven't observed that to be the case. I understand that some data, like the Teams usage report, have caveats stating "To ensure data quality, we perform daily data validation checks for the past three days and will be filling any gaps detected. You may notice differences in historical data during the process.” but we are able to get data the following morning for previous days. Either way, time delays for previous day(s) aside, I'm still primarily concerned with clarifying the begin date time for the 30 day report (as opposed to the end date time), with emphasis on the "time" portion of the date time.
To clarify my question, does the 30 day report look at data from exactly 720 hours from the moment it was queried, or does it look as far back as midnight on the calendar date that is 720 hours from the moment it was queried? I suspect the former, but am trying to confirm.
![](data:image/svg+xml, %3Csvg xmlns='http://www.w3.org/2000/svg' height='64' class='font-weight-bold' style='font: 600 30.11764705882353px "SegoeUI", Arial' width='64'%3E%3Ccircle fill='hsl(0, 96%, 10%)' cx='32' cy='32' r='32' /%3E%3Ctext x='50%25' y='55%25' dominant-baseline='middle' text-anchor='middle' fill='%23FFF' %3EYH%3C/text%3E%3C/svg%3E)
Hello ShaSap,
Thank you for reaching out to Microsoft Support!
Extract 30 days as the first 30 days from your current time, if your current time is October 1, then the first 30 days is September 1 - October 1 30 days, according to the document, the date format is YYYY-MM-DD, the time is accurate to days, not accurate to seconds, and when just tested, the endpoint's current range is 28 days.
Hope this helps.
If the answer is helpful, please click Accept Answer and kindly upvote it. If you have any further questions about this answer, please click Comment.
Hi Yakun. Thanks for your response. Unfortunately, since you don't explicitly state what timeframe the "30 days" entails, I'm afraid I'm still unclear. If (in the example) the api is including anything with a date of 30Sep-01Oct (inclusive), then the applicable date time frame any data between 30-Sep 00:00:00 UTC to 01Oct 23:59:59 UTC, but you don't explicitly state whether that is or is not the case; simply saying "September 1 - October 1" does not provide the level of fidelity that I'm searching for.
To ask my question in another form: Is the api including the past 30 days (exactly 720 hours) or anything within the past 30 calendar days (which would be more than 720 hours if run after midnight)? I assume it is the former, but am trying to confirm that.
It should be anything in the last 30 days, even if it runs after midnight, more than 720 hours, because according to the documentation, the event is accurate to the day, so if we do the query at midnight and fill in the time just before the 30th day, according to the documentation, this should be searchable and you can verify it.
The most detail I can find in online documentation are the function parameters for the HTTP request for the various reports, but it still doesn't describe what "D30" (30 days) entails. Documentation for each report itself is also vague by stating "trends over the last 7 days, 30 days, 90 days, or 180 days." If there another documentation you've found that describes in more detail how the report is generated I'd love to check it out!