Re: [Debconf-team] Measuring DebConf success (Was: Regarding DebConf13 planned location)
On 27/11/12 19:42, Andreas Tille wrote:
>
> What I want to say is: There is neither any objective nor most probably
> any subjective measure to declare a DebConf a success or failure. The
> only thing we can really measure is the surplus / dept a DebConf
> created. I also would not fully agree that this number is correlated
>
Andreas, I think this is an excellent topic and it is not just
applicable to the DebConf13 discussion
For the question of `success' or `failure', we would first need to
define outright objectives. I believe these already exist and I saw
them on a slide presented at DebConf12. If a DebConf meets those
objectives, it would be considered a success.
http://penta.debconf.org/dc12_schedule/events/905.en.html (40 seconds into the video)
"Primary goals:
- enable face-to-face interactions
- provide talks and video
- provide time to work on Debian
Secondary goals:
- motivate contributors
- motivate the local community"
I agree it is hard to quantify the success with a `score', but if those
goals were used, the score could be calculated based on each of them:
A: enable face-to-face interactions: how many people attended?
B: provide talks and video: how many hours of video created?
C: provide time to work on Debian: how many people * how many hours?
score = A + B + C ???
and then if money is relevant to the discussion, the score could be
divided by the cost of the event.
Rather than talking about it in absolute `scores', it may be meaningful
to ask questions like:
a) how did the (cost adjusted) score compare to the previous DebConf in
the same region?
b) how did the (cost adjusted) score compare to a mini-DebConf or other
low cost/less effort gathering?
c) in qualitative terms, how did it compare to other open source events
in the same region?
If we find that the score of a mini-DebConf is massively bigger, then it
would suggest it is better to run more of them than one big DebConf - or
we need new objectives.
Reply to: