toggle quoted messageShow quoted text
I would add that your marketing/communication department probably already has an analytics platform that is cleared by all the necessary people in the organization and if possible, you should try and piggy back on that setup.
If GA isn't an option (for the multitude of reasons mentioned earlier), there is an open source, but very competent option called Mamoto, https://matomo.org/
, which should be easy for your IT group to setup and run.
On 10/01/2018 11:30 AM, Tony Chung @codemaestro
I echo the concerns in Mark's and Nicholas' messages that your secure site might not be exposed to Google's API. Also, your clients might not take too kindly to their usage tracking data being stored on Google's servers, albeit mostly irrelevant outside of your company's needs.
You could review the log files, but that could require a lot more data massaging to get the specific information you want. Plus, server logs only show request access. They won't give the keywords or other rich data that will help you to improve your docs.
You might want to implement an analytics tool that captures and stores data directly onto your own servers. You should still get buy-in from your customers that you will be collecting this data, but assure them its for your own proprietary use, to improve quality, and the information will not be shared.
If you have tech-savvy IT people (who understand the web, not just network infrastructure and hardware/software), you could implement Matomo (fomerly Piwik free) https://matomo.org/ . For more guidance, you can pay for Piwik Pro to have them work it up for you.
On Mon, Oct 1, 2018 at 11:05 AM Mark Giffin mark@... <mailto:mark@...> [dita-users] <dita-users@... <mailto:dita-users@...>> wrote:
Google Analytics is probably the easiest way to find out what pages
are being accessed. However I'm not sure how GA works with a secure
site that is not exposed to the internet. Maybe it just works, but
the GA code you put in your HTML pages needs to talk to the Google
borg in some way, so they can generate reports pages and so on.
But if you can get access to your web server logs from the web
server that hosts your webhelp, you can get the same information on
pages accessed. Plus with the web logs you can probably figure out
the search terms being searched for. This would be more work than
using GA, but it can probably give you more info.
By the way I assume by "search" you mean the full-text search that
gets generated with Oxygen webhelp. But since this is an internal
help system, maybe you are using some other search platform like
Solr, Elasticsearch or Lucene. In that case, those tools have ways
to tell you what terms are being searched for.
Mark Giffin Consulting, Inc.
On 10/1/2018 8:10 AM, Nicholas Mucks urbanrobots@...
<mailto:urbanrobots@...> [dita-users] wrote:
We gather analytics on some of our webhelp outputs. You’ll need to
set up your GA account and include a script tag in the header or
footer of each page that points to one of your account’s tracking IDs.
By the way, all, we’ve had debates about the ethics of capturing
user traffic with GA. We have some people who are against
gathering this information on privacy grounds and others who push
for GA in order to improve the docs. This becomes especially
sensitive when developing docs delivered internationally.
Sent from mobile
On Oct 1, 2018, at 1:55 AM, Radu Coravu radu_coravu@...
For the Oxygen user's manual which is produced also with WebHelp
responsive we add to each HTML page a custom footer with a
I'm not sure if we have an API to give you access to the searched
Maybe you can write to "support@...
<mailto:support@...>" about this use case and
if we do not have support for it we could add some in a future
<oXygen/> XML Editor
On 10/1/2018 11:18 AM, ann.jensen74@...
<mailto:ann.jensen74@...> [dita-users] wrote:
> We develop our product documentation in DITA (Oxygen XML
> Author), publish using Oxygen WebHelp and deploy it to a secure
> (only accessible via our web-based product). We would like to
> information on what pages our users are accessing and what keywords
> people are searching for.
> Is anyone else managing to gather this information from WebHelp
> published content for the purpose reporting and analytics?
> Any feedback appreciated,