Statistics on number of forum threads answered
Jason Barnabe (np)
up being answered or not. These numbers are based on a manual count of
threads last posted in on Thursday, February 28. There were a total of
166 threads sampled. I've categorized them as follows:
A) No contributor response. The end user did not receive an answer
from us, and has given no indication that they've solved it on their
own. These threads should be addressed by contributors. 20 instances
for 12% of threads.
B) Contributor response, and no reply from end user. Contributors gave
a possible solution to the end user's problem, but the end user didn't
answer to say whether it worked or not, or didn't give further
information requested by the contributor. These threads can be
addressed by contributors (to provide further answers), but they
likely need a end user to move forward. 89 instances for 54% of
threads.
C) Contributor and end user discussion, but no definitive answer. The
end user and contributors posted a few messages back and forth. These
threads should probably be addressed by contributors. 23 instances for
14% of threads.
D) Problem solved. The end user says their problem is solved. No
further action is required from the contributor to help the user. 23
instances for 14% of threads.
E) Miscellaneous. Duplicates, posts about Thunderbird, etc. No action
required from contributors. 11 instances for 7%.
Some analysis:
Only 26% of threads (A + C) look like they require further contributor
posts. It's good that it's such a small proportion, but this makes it
difficult to find people to help. If contributors have a visible
indication of threads that are answered (bug 397703), then this
increases to 30%. If contributors also have an indication that the
last poster in a thread was a contributor, then this increases to 80%.
In the majority of threads, the end user doesn't post back. I'm
worried that a significant number of these are people who can't find
their thread or don't check back. We default registered users to get
notified, but most end users aren't registered. Bug 398486 would help
with this.
A significant number of users are still posting about Thunderbird. I'm
not sure what else we can do to prevent this, but it would be useful
to have some text that's easy to put in directing these users to an
appropriate place.
bba...@gmail.com
It's pretty cool to see some numbers; I hope you'll continue running
statistics periodically.
Comments:
I'm surprised (A) is above 10%. There are some threads I tend to leave
alone, hoping someone else will be able to help.
These include:
(1) Posts that deal issues external to Firefox on Macs (I don't have a
Mac, and can't test any fixes)
(2) Ditto for flavors of Linux I don't have, plus I implicitly assume
if you're smart enough to be running an esoteric version of Linux,
more good can be done in a scarce amount of time answering 0 reply
threads for more helpless users.
(3) Posts where I can't make out what the problem is, due to poor
writing
(4) Posts where the poster claims to have tried everything I would
have recommended
(5) Posts where the poster is not friendly, threatening to quit using
Firefox, and the like
If we could get a Mac expert to hang around, that would help with (1).
Similar comments for (2) - my assumption may be wrong, but I think my
reasoning about maximizing help per time stands.
I'll make it a point to ask for clarification in the future for (3).
I'm not sure what to do about (4), especially when it's combined with
(5).
I usually ignore unfriendly components and just post like I would have
anyway, but when my time is scarce, I'll gladly help a friendly poster
over an unfriendly one.
(B) is no surprise. I agree on Bug 398486 helping this lots.
I always feel bad about (C), if a thread has dropped off my radar and
the user posts back asking again.
I used to only hit "Watch this thread" for posters who present with
interesting problems where I think I might learn a new trick, but
I've started hitting it more often for more diagnostically boring
problems.
When I'm looking through the forum, I usually read posts with a non-
contributor on the last post, but with scarce time, you can't catch
them all. There's also the factor of scarce time and points (2), (4),
and (5) above.
I think (C) could be helped by some sort of way of indicating to other
contributors that one contributor has seen the thread, but doesn't
know how to solve it.
For example, let's say there's a thread about not being able to access
a site on a Mac. I post the usual clear cookies/cache routine, but the
user comes back and says it works in Safari, but gives some obscure
Apple error message. If I could mark that thread "S.O.S." so others
logged in as contributors could see it still needs attention, I think
that would help. This would be distinct from the normal "Thread
unsolved"/"Thread solved" indicator.
Good stuff; I hope we can improve the stats!