Hi everyone,
I've only posted once before here, and didn't do much of an intro back then, so let me do one now. I'm the Program Manager for General Engineering at Wikimedia Foundation, which is the slice of the WMF Engineering organization that does infrastructure-related software development. One piece we're responsible for is the analytics infrastructure.
We're in the process of planning our software development for analytics for the coming months, so we've had a few conversations, and Howie Fung and I spent some time planning and writing up our thoughts on feature prioritization here: http://strategy.wikimedia.org/wiki/Task_force/Analytics/Feature_prioritizati...
This is a really rough cut, and something we haven't fully discussed within the Foundation, so don't take this as something that is coming down from on high. There are some things on the list that are well underway, but many things are things we're just getting started on.
Barring any objections here, we'd like to use this mailing list as our primary venue for discussing general prioritization of analytics features. We know we need a place that we can tell WMF employees to subscribe if they're interested in this stuff, and nothing we're discussing should be confidential. Rather than starting a new mailing list, we'd like to try using this list for a bit (in combination with relevant talk pages on documents referenced here). If it turns out we're generating enough traffic to warrant splitting off or if this list isn't working out for whatever reason, we'll figure out some alternate plan.
While we suspect that many of the details will be of specific interest to Foundation employees (who are relying on much of this information to perform their jobs effectively), we also know there is plenty of general interest in this work. Please feel free to share your thoughts.
Thanks! Rob
Hi Rob,
For one, I'm super pleased that we're taking a wholistic approach to improving the analytics on WMF project. I have been hoping that we make it easier to extract x, y or z stats/metrics on an ad hoc basis, but to actually get proper analytics built right in is a giant leap beyond what I thought was possible. And secondly, as far as I'm personally concerned, this research-l mailing list would seem an appropriate place to host discussions about the analytics project in the manner in which you described.
One question: as I understand it, one of the key priorities of this analytics project is the installation of OpenWebAnalytics (which AFAICT will be similar to GoogleAnalytics but open source and also compliant with the WMF's stringent privacy policy). If so, will the full array of anonymised analytics be visible to everyone live, or will the results be released in a summarised format on a regular basis? That is, will the public/wikimedians/press be able to see the same thing that the WMF can see and at the same time?
Finally, if I may just throw in a little request to the "wishlist" - one thing that GLAM partners would really like to be able to do is easily produce for themselves a "report card" of their organisation's relationship to Wikimedia over time. Currently, we make do with producing ad hoc stats for them based maingly on magnus' tools (especially baGLAMa and GLAMorous) and other things like linkypedia.inkdroid.org . It would be brilliant if a GLAM partner could quickly and easily produce a *pretty* report that showed how their images were being used (number of usages, number of views...), how our external links to their site were used (most popular referral paths, total traffic, most linked-from categories...) and how articles about things relate to them are used (quality improvement over time, combined pageviews for categories important to them...). Ideally, if this could generate into a report fit to show to senior management, I suspect that we would have much greater success with enticing more GLAMs to move towards free-culture. All "whishlist" stuff I know, but I thought I might as well ask :-)
-Liam / witty lama
On 07/01/2011, Rob Lanphier robla@wikimedia.org wrote:
Hi everyone,
I've only posted once before here, and didn't do much of an intro back then, so let me do one now. I'm the Program Manager for General Engineering at Wikimedia Foundation, which is the slice of the WMF Engineering organization that does infrastructure-related software development. One piece we're responsible for is the analytics infrastructure.
We're in the process of planning our software development for analytics for the coming months, so we've had a few conversations, and Howie Fung and I spent some time planning and writing up our thoughts on feature prioritization here: http://strategy.wikimedia.org/wiki/Task_force/Analytics/Feature_prioritizati...
This is a really rough cut, and something we haven't fully discussed within the Foundation, so don't take this as something that is coming down from on high. There are some things on the list that are well underway, but many things are things we're just getting started on.
Barring any objections here, we'd like to use this mailing list as our primary venue for discussing general prioritization of analytics features. We know we need a place that we can tell WMF employees to subscribe if they're interested in this stuff, and nothing we're discussing should be confidential. Rather than starting a new mailing list, we'd like to try using this list for a bit (in combination with relevant talk pages on documents referenced here). If it turns out we're generating enough traffic to warrant splitting off or if this list isn't working out for whatever reason, we'll figure out some alternate plan.
While we suspect that many of the details will be of specific interest to Foundation employees (who are relying on much of this information to perform their jobs effectively), we also know there is plenty of general interest in this work. Please feel free to share your thoughts.
Thanks! Rob
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
Hoi, In order to celebrate 10 years and prepare for the next 10 years, there will be a hackaton in Amsterdam. We will concentrate on GLAM stuff. Yes, we hope that Erik Zachte will have time to come as well, sadly doubtful, and one of the things high on the list is to streamline some of Magnus' wonderful tools. Thanks, GerardM
On 10 January 2011 05:35, Liam Wyatt liamwyatt@gmail.com wrote:
Hi Rob,
For one, I'm super pleased that we're taking a wholistic approach to improving the analytics on WMF project. I have been hoping that we make it easier to extract x, y or z stats/metrics on an ad hoc basis, but to actually get proper analytics built right in is a giant leap beyond what I thought was possible. And secondly, as far as I'm personally concerned, this research-l mailing list would seem an appropriate place to host discussions about the analytics project in the manner in which you described.
One question: as I understand it, one of the key priorities of this analytics project is the installation of OpenWebAnalytics (which AFAICT will be similar to GoogleAnalytics but open source and also compliant with the WMF's stringent privacy policy). If so, will the full array of anonymised analytics be visible to everyone live, or will the results be released in a summarised format on a regular basis? That is, will the public/wikimedians/press be able to see the same thing that the WMF can see and at the same time?
Finally, if I may just throw in a little request to the "wishlist" - one thing that GLAM partners would really like to be able to do is easily produce for themselves a "report card" of their organisation's relationship to Wikimedia over time. Currently, we make do with producing ad hoc stats for them based maingly on magnus' tools (especially baGLAMa and GLAMorous) and other things like linkypedia.inkdroid.org . It would be brilliant if a GLAM partner could quickly and easily produce a *pretty* report that showed how their images were being used (number of usages, number of views...), how our external links to their site were used (most popular referral paths, total traffic, most linked-from categories...) and how articles about things relate to them are used (quality improvement over time, combined pageviews for categories important to them...). Ideally, if this could generate into a report fit to show to senior management, I suspect that we would have much greater success with enticing more GLAMs to move towards free-culture. All "whishlist" stuff I know, but I thought I might as well ask :-)
-Liam / witty lama
On 07/01/2011, Rob Lanphier robla@wikimedia.org wrote:
Hi everyone,
I've only posted once before here, and didn't do much of an intro back then, so let me do one now. I'm the Program Manager for General Engineering at Wikimedia Foundation, which is the slice of the WMF Engineering organization that does infrastructure-related software development. One piece we're responsible for is the analytics infrastructure.
We're in the process of planning our software development for analytics for the coming months, so we've had a few conversations, and Howie Fung and I spent some time planning and writing up our thoughts on feature prioritization here:
http://strategy.wikimedia.org/wiki/Task_force/Analytics/Feature_prioritizati...
This is a really rough cut, and something we haven't fully discussed within the Foundation, so don't take this as something that is coming down from on high. There are some things on the list that are well underway, but many things are things we're just getting started on.
Barring any objections here, we'd like to use this mailing list as our primary venue for discussing general prioritization of analytics features. We know we need a place that we can tell WMF employees to subscribe if they're interested in this stuff, and nothing we're discussing should be confidential. Rather than starting a new mailing list, we'd like to try using this list for a bit (in combination with relevant talk pages on documents referenced here). If it turns out we're generating enough traffic to warrant splitting off or if this list isn't working out for whatever reason, we'll figure out some alternate plan.
While we suspect that many of the details will be of specific interest to Foundation employees (who are relying on much of this information to perform their jobs effectively), we also know there is plenty of general interest in this work. Please feel free to share your thoughts.
Thanks! Rob
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
-- wittylama.com/blog Peace, love & metadata
Wiki-research-l mailing list Wiki-research-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
On Sun, Jan 9, 2011 at 8:35 PM, Liam Wyatt liamwyatt@gmail.com wrote:
For one, I'm super pleased that we're taking a wholistic approach to improving the analytics on WMF project. I have been hoping that we make it easier to extract x, y or z stats/metrics on an ad hoc basis, but to actually get proper analytics built right in is a giant leap beyond what I thought was possible. And secondly, as far as I'm personally concerned, this research-l mailing list would seem an appropriate place to host discussions about the analytics project in the manner in which you described.
Excellent! There's some wording on this page that caused me to be a little timid about this: https://lists.wikimedia.org/mailman/listinfo/wiki-research-l
"Internal Wikimedia matters, discussions of new projects and similar threads should be kept off the list."
This is arguably an "internal Wikimedia matter", but I suspect that wording was written long ago, and could use some tuning and clarification.
One question: as I understand it, one of the key priorities of this analytics project is the installation of OpenWebAnalytics (which AFAICT will be similar to GoogleAnalytics but open source and also compliant with the WMF's stringent privacy policy). If so, will the full array of anonymised analytics be visible to everyone live, or will the results be released in a summarised format on a regular basis? That is, will the public/wikimedians/press be able to see the same thing that the WMF can see and at the same time?
Not yet. We've discussed how to make this possible, but I think there's a lot of work left to do to make this a reality. We'd need to make sure of a couple of things: 1. That the only thing we're providing is a fully sanitized view of the data 2. That any user interface that we expose via public web page go through much more rigorous security review
For the first item, it's worth discussing on one of the OWA mailing lists: http://wiki.openwebanalytics.com/index.php?title=Support
I'll also make Peter aware of this thread so he knows what's going on.
Finally, if I may just throw in a little request to the "wishlist" - one thing that GLAM partners would really like to be able to do is easily produce for themselves a "report card" of their organisation's relationship to Wikimedia over time. Currently, we make do with producing ad hoc stats for them based maingly on magnus' tools (especially baGLAMa and GLAMorous) and other things like linkypedia.inkdroid.org . It would be brilliant if a GLAM partner could quickly and easily produce a *pretty* report that showed how their images were being used (number of usages, number of views...), how our external links to their site were used (most popular referral paths, total traffic, most linked-from categories...) and how articles about things relate to them are used (quality improvement over time, combined pageviews for categories important to them...). Ideally, if this could generate into a report fit to show to senior management, I suspect that we would have much greater success with enticing more GLAMs to move towards free-culture. All "whishlist" stuff I know, but I thought I might as well ask :-)
By all means. I'm wondering what the most sensible way to organize and vet all of the community wishlist issues. What I'd like to do is make sure we have a bulleted summary or a query somewhere that we can march through during the meetings we have at WMF about priority setting. If it's buried in an email thread, it's going to get lost. Where do you think is the most sensible place to ask people to put these requests that works for everyone?
Rob
wiki-research-l@lists.wikimedia.org