Google Analytics
Apr. 20th, 2009 01:53 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
foxfirefey posting in ![[community profile]](https://www.dreamwidth.org/img/silk/identity/community.png){kind=small}
dreamwidth_metaLiveJournal has never allowed any real web analytics to be added to personal journals, although sponsored communities were able to get them. Sure, you could add stat counters or web bugs from LJ Toys. But I'm unaware of any way on LiveJournal to get the referral URL of people who were linking to your post, save for the recently implemented and entirely optional pingbacks.
Dreamwidth, however, is going to give paid users Google Analytics as a feature. This means that paid users will be able to know who in DW is linking to them, leading to some interesting changes from the way things used to be. I think this has the potential to surprise and upset people.
For instance, let's say you link to someone's post in a friends only post in your journal or use <user name="user"> to link to their journal in a locked post. Some of your access given subscribers click on that link, and if the user you linked to is paid and using Google Analytics, they'll know you were talking about them in a post they don't have access to, and if you linked to a specific post, they'll know which post you're talking about. Stealth talking about people has become that much harder and unreliable.
There's a limited ability to avoid this. URLs are automatically turned into links; you can do formatting to make it unlinked, so people have to copy and paste, but some people have browser extensions that will autolink anything that looks close to a URL, so you can't always depend on that. You'll have to go above and beyond to obfuscate the link to make sure that doesn't happen and not use user tags to link to someone--but if you don't do that, someone is bound to make a Greasemonkey script that could go to a highlighted name, and they'll still get the referral. Edit:charmian and kaki point out that URL obfuscators might get used more, like TinyURL and anonym.to. I agree with this! However, there are even browse add ons that resolve those services to their actual URLs, so even that is not a failsafe.
What effects do you think this is going to have on social interactions on Dreamwidth? What other effects will Google Analytics have on users?
Dreamwidth, however, is going to give paid users Google Analytics as a feature. This means that paid users will be able to know who in DW is linking to them, leading to some interesting changes from the way things used to be. I think this has the potential to surprise and upset people.
For instance, let's say you link to someone's post in a friends only post in your journal or use <user name="user"> to link to their journal in a locked post. Some of your access given subscribers click on that link, and if the user you linked to is paid and using Google Analytics, they'll know you were talking about them in a post they don't have access to, and if you linked to a specific post, they'll know which post you're talking about. Stealth talking about people has become that much harder and unreliable.
There's a limited ability to avoid this. URLs are automatically turned into links; you can do formatting to make it unlinked, so people have to copy and paste, but some people have browser extensions that will autolink anything that looks close to a URL, so you can't always depend on that. You'll have to go above and beyond to obfuscate the link to make sure that doesn't happen and not use user tags to link to someone--but if you don't do that, someone is bound to make a Greasemonkey script that could go to a highlighted name, and they'll still get the referral. Edit:
charmian and kaki point out that URL obfuscators might get used more, like TinyURL and anonym.to. I agree with this! However, there are even browse add ons that resolve those services to their actual URLs, so even that is not a failsafe.What effects do you think this is going to have on social interactions on Dreamwidth? What other effects will Google Analytics have on users?
no subject
Date: 2009-04-21 12:21 am (UTC)I think the key is partly technological choice and transparency, partly adjusting social norms (those who are used to blogs will probably be less freaked out by this change, esp. when you consider those widgets that display on the front of blogs where visitors are coming from are much more public), and partly individual management (people knowing what they are comfortable knowing or having known about them and how to manage that).
no subject
Date: 2009-04-21 12:32 am (UTC)To be honest, I think that clearly identifying who at the current time has GA on is deceptive. Somebody could set up or lose Google Analytics at any point in time. There's no good way to warn people of who has GA enabled *before* they go to their journal, either. The best way would be to put a notification on the profile, but I think the profile itself gets to have Google Analytics on it (could be wrong) and so that kind of defeats the whole purpose. But, we do know at the very least that only paid users can, and they won't be a majority of users.
Google Analytics is easy for an individual to block--that is, somebody who doesn't want to be counted in someone else's stats can easily install AdBlockPlus and block Google Analytics.
no subject
Date: 2009-04-21 02:58 am (UTC)And I see what you mean about identifying who has GA or not. I put LJ Toys on my posts for about a week once and then took it off again (except my profile, oops! *fixes*). Besides the weirdness of adjusting to knowing who was hitting what (except those who had it blocked), it was eye-opening how much general info I got. I felt weird about not letting people know but after that week any autonotification that I had LJ Toys would have been wrong.
And it's good to know AdBlockPlus can take care of that.
no subject
Date: 2009-04-22 08:49 pm (UTC)no subject
Date: 2009-04-22 08:51 pm (UTC)