[rt-users] What RT statistics/metrics do you report on?

[Kenneth Crocker]

Steve,

I believe the "Time Worked" report offered in the wiki has some of the
answers for you. That code pulls up time reported from the TRANSACTION
table. So whether or not a ticket is resolved bears doesn't matter, neither
does the ticket owner. Anyone who reports time against a ticket creates a
record in t he transaction file. Sooo, you might want to look at that code
and borrow some of it for your own purposes.

Kenn
LBNL

On Tue, Jul 5, 2011 at 9:07 AM, Hersker, Steve <SHersker at tngus.com> wrote:

> Not a technical issue, but I was wondering what RT stats folks report
> on/review periodically?
>
> For 5+ years, our RT reports consisted of exporting to an Excel
> spreadsheet followed by copy/paste/update of formulas/etc.  Mildly time
> consuming but the larger issues we had were that we were unable to
> report stats based on a user's Organization and we were unable to get a
> lot of detail on a per-engineer basis.
>
> When we upgraded to RT 4.0, I took the opportunity to add a couple of
> custom fields and wrote a Perl script to extract and report on various
> metrics instead of using Excel. This allows us to now report on
> Division/Site (stored in the Organization field) and get more detail for
> each support engineer.
>
> For example, we report monthly on the following (both overall and per
> support engineer):
> Tickets Created
> Tickets Resolved
> Hours recorded & % of hours recorded (based on business days in month)
> Average minutes per ticket
> % Submitted via Phone
> % Submitted via Email
> Average days to resolve
> Resolved Same Day
> Resolved Next Day
> Resolved two or more days
>  [A note on the "Resolved" calculations: I use date::manip to calculate
> business days instead of calendar days. Eg a ticket comes in at 4:30pm
> on Friday and is resolved by 9:00am Monday, that's "resolved - same day"
> - within 8 business hours. The old Excel reports would show that as
> resolved in 3 days, which to me wasn't fair.]
>
> Ticket count, hours and Avg min/ticket (% and total for each) for
> - Ticket "Type" (Issue, Request or Project)
> - Ticket "Category" (problem category and topic, eg. "Workstation" and
> then within that "Application install/fix", "hardware repair/upgrade",
> "OS fix/rebuild", etc)
> - Division
>
> Then, "Top Five" by both ticket count and time recorded:
> - Category / Topic
> - Site
>
> One small issue is that ticket time is only allocated to the Owner of
> the ticket. So, if someone records 2 hours of time on a ticket they
> don't own, they don't get credit for it. But, in our group that's a
> fairly small occurrence. Also "time recorded" is only reported on
> tickets that were Resolved the previous month. So, if an engineer logged
> 40 hours on a ticket but didn't resolve it yet, that time doesn't show
> up until the next month (or whenever they resolve the ticket). I'd
> rather see actual time recorded in a month but, again, not yet worth the
> effort to adjust the report, especially since the time shows up sooner
> or later.
>
> One stat I'd like to add soon is Average Minutes per User (per Site).
> For example, if we have Site A with 50 users and Site B with 5 users and
> both have 10 hours of time, that might not stand out in the "Top Five"
> but would in Minutes/User, prompting me to wonder and research why we
> spend 10x more support for the smaller site etc.
>
> I realize with RT's infinite flexibility, there are a great many ways to
> use it. But, I am curious to see what else anyone might report on?
>
> Thanks!
> Steve
>
>
>
> --------
> 2011 Training: http://bestpractical.com/services/training.html
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.bestpractical.com/pipermail/rt-users/attachments/20110705/c915089d/attachment.htm>