my customer asked me for a web controlling tool. because i’m not really
fan of hosting all requestdata at google-servers i want to install a tool
on the webserver and track userrequests there. i tested different tools
and my favorite has been trace watch, because i like the pathanalysis-feature. now i’m testing piwik, because
it seems to be the best choice for the future. the modular concept makes
a good impression but the only thing i miss is an information about what
the users klicked to reach desired subsite.
is there such a feature and i didn’t found it yet ?
thanks for tipps und
Nice to see that development on Tracewatch has resumed after a 4-5 year hiatus.
Click stream/click path analysis is not on the Piwik 1.0 roadmap. This is a frequently requested feature and we’d love to have it, so if anyone wants to help out…
If i could do a simple HTML page that simulate the view part of the MCV pattern (i’ve been reading about Zend), could that help ?
As a first small step i mean.
I know main Piwik devs wont work on this for now but maybe it could help giving someone else the interest in helping.
What i’m not sure yet is how to use the Piwik CSS to have the same look and small other details like that.
And a few hours later, here it is :
[size=8pt]Copy the URL as this forum dont permit to post links.[/size]
It show fake data. It use some of Piwik CSS and images (that i’ve duplicated in a new folder on my site). The JS used to show more URLs has been done by Dream CS4; i got lazy. But i still had to modify it so it change the display instead of the visibility.
Ideally there would be a way to refold the URL list after unfolding it but i didn’t make that part.
I put the full (fakes) IPs in there but for German people i gess it would be anonymized a bit.
Edit : i will leave that page in place as long as it will be usefull.
Sorry to be bothering, but has anyone visited the page i made ?
I didn’t put any tracker code in it.
Anyone has any comments ?
I like the mock-up.
Do you have any ideas for an aggregated report?
I’m sorry, maybe because english is not my first language, i’m not sure what you mean by “aggregated report”.
It’s not the export to CSV or anything like that ?
Thanks, for looking and commenting, by the way.
aggregate means something like “to put together”, something like an overview.
But I´m not sure if a click path tracking needs something aggregated. Each click path is unique. Youre example is what I expected for a path tracking.
When you consider the use of the back button and open-in-a-new-tab, you have to change your thinking of a click path. It isn’t necessarily a straight-line, A to B to C. It could be A to B and A to C, or circuitous, A to B to A to C.
I like your mockup and linked it from: http://dev.piwik.org/trac/ticket/1120
if you can propose a mockup that answers all the points in the ticket, that would be ideal. Can you look into proposing a patch to the existing templates? Thanks!
Oh, but Piwik doesn’t track from wich page another has been accessed, no ?
That would mean for each action to save in DB on wich page it has been perfomed each time.
That could mean for every page on a given site there would be like 10 or 20 differents actions if the page can be accessed from 10 to 20 others pages.
Same thing for outlinks and download.
But if it’s really possible to do that, we simply have to use UL element inside one another like this :
Erf! It's doesn't work good in the forum; i'll make a second table on the same page as the first one.
EDIT : My “mockup” page has been updated now.
We don’t store the referer for each page view indeed, this would be too much data and not useful enough for the analysis to justify the cost. TulipVorax, if you want your code considered or the plugin Live!, please comment in the ticket and try to implement the bug fixes/improvemenst in the ticket, this would help changing for your version. Cheers
I understand the vastly expanding DB demands with such plug-in, but it could be something off by default with the warning.
I think Google’s version works nicely: you can track path of individual user, or on page by page basis: on each specific page there is a breakdown in where the visitor came from (direct entry (so many clicks), search engine (), page A (), page B () etc.), and where the visitor went (page C (), page A (), exit () etc.).
we have just released a new feature, “Page Transitions” which shows the path of a visitor from a given page, and where visitors click from there. It’s pretty awesome! Check it out: Clickpath analytics