Because HBR really needs more "how should the CEO fix this PR meltdown?" case studies
Ah, what a lovely week to have built a social networking site based on tons of people writing whatever they want, and then try to co-opt the "whatever they want" part. Sow the wind, and all that.
We'll start with the newest fashion, Facebook, where the slow burn over their new Beacon "service" pretty much boiled over onto the stove this week. For those who haven't heard, this is a cooperative bit of javascripting with some other sites like Travelocity, Blockbuster and Overstock to automatically update your Facebook News Feed when you do something (like make a purchase) on the partner site - unless you spot the item-by-item opt out and click it quickly before the update happens.
Funny thing, people actually objected to having their surfing habits and, in some cases, online purchases collected and posted for others without being asked to opt-in or turn it on first and with only limited ability to opt-out on an ongoing basis. Go figure! Aside from the "you published my hemorrhoid cream purchase where?" concerns, what could possibly go wrong with publishing your recent online purchases to all your friends starting, with little notice, in mid-November?
Chad Stoller, of the online advertising agency Organic, demonstrated that it's apparently possible to get a highly-placed position in the PR and marketing field despite showing the sort of judgement that normally gets the pros telling you to shut up and stay away from the press before you screw things up even further by getting quoted it in the Times "poking fun" at Facebook users, many of whom discuss odd personal details online, by saying “Isn’t this community getting a little hypocritical? Now, all of a sudden, they don’t want to share something?”
Facebook is now backpedaling, although there are some contradictions in the reports about how things will change, and some of the partners have apparently backed out until they see more opt-in emphasis. In the meantime, if you're using Firefox, there are instructions for using a site blocking download to automatically block Beacon traffic - until and unless Facebook decides getting Beacon stats whether you approve or not is important enough to make trivial changes to the Beacon collection URL...
Next, on to this week's joys with LJ and Kid-Friendliness. The official details are on lj_biz (with a great deal of user response, some of which is the only place where important details have been documented so far) and the FAQ, with some completely unoffical highlight pointers on the_lj_herald, but this is what it looks like:
The devil will quickly be in the details, but we'll start with the quick form, or "what it looks like they were trying to do" and drill in as we go. First, they've introduced the ability to assign adult-content status to your own posts. There are two new levels other than "ok for everyone" which are "Adult Concepts" and "Explicit Adult Content". They're pretty vague and subjective in their definitions - we'll come back to that - but the idea is they'll then make some attempt to keep people under age 14 from seeing the former, and people under 18 from seeing the latter. There's room for some eye rolling, especially as you get into the implementation, but giving you one more way to slap metadata on your journal and postings is not the part getting the lion's share of the controversy.
Second, they've introduced the ability to assign those levels to your entire personal or community journal. That level then acts as the default for any postings to the journal, and for community journals you can't "join" the community if you are the wrong age. The obvious question is whether you're supposed to mark your journal if (to take the extremes) it "may some day contain a post of that sort" or "all of its posts are of that sort". Looking at what the LJ posts actually say as you dig into the comment responses, they mean the latter - the idea is if the nature of the journal or community is that you're usually going to be using one of the restrictions (they "suggest" if it's 50% or more) that you can just set it once as a journal-wide default. That distinction has been very poorly described in the announcements (such as even they are) - many people think if they have ever posted (or may ever post) something that qualifies, that they now "have" to mark their journal just in case. Of course, many people are marking their journals as "adult content" just as a protest, but at least some are doing it because the distinction wasn't originally made clear. That, in my opinion, is the biggest source of sturm und drang for them now on this specific point - the poorly-executed launch, rather than the actual problems in how it works.
Third, we open the really big can of worms: not everyone is going to volunteer to use this, which could mean kids still seeing things that would frighten the horses. Soooo... there's now an option to narc out someone else's postings (or entire journals) as Explicit Adult Content. This is the part that immediately raises a whole lot more questions, which they somehow didn't realize were important enough to lay out answers to until they started getting flooded with negative responses.
Most of those questions are addressed, at least in part, by spelling out what actually happens when a reader presses the "flag this" button on someone else, which wasn't fully explained until the follow-ups. What does happen? I go to your journal, say, see naked people, and press the "flag this" button. I get prompted for several categories of objection, some of which dump directly into existing "maybe this needs to get taken down and the poster banned" customer service pipelines and really have little to do with the new function. But one is "this is explicit adult content" which is our new label, and I click that. But there's only one of me; so far, nothing actually happens. Several other folks come along and also so flag it. Nothing (immediately or automatically) happens as far as the outside world can see, but it now goes into a queue to have an LJ moderator look at it and make a judgement call because more than one reader has weighed in on it. If the LJ employees agree, then it gets marked explicit adult content whether the original poster wants it to or not. At this point, we probably just want to go straight to bullet points.
- Doesn't the vagueness and subjectiveness of the definitions pose a problem? Yep; the fact that the decision ultimately gets made by a (relatively) small group of LJ employees who are comparing notes may help make the calls more consistent and not least-common-denominator (your optimisim about that may vary) but subjective calls are the sort of black hole you really want to think long and hard about signing up for when you run a site like this.
- Doesn't this mean that LJ is going to have to have a whole mess of people to make these judgement calls? Well, yes, although they've already signed on for that particular ride based on their existing policies for reporting offensive or illegal content, so it looks like they're gambling that this will mostly pre-categorize the tattling that they're already getting so they can at least triage "this is sort of adultish" from "this is blatantly illegal kiddee porn" more than it creates additional moderation burden. Time will tell if they're right about that. The fact that they make the "flag this" button big, colorful, and easy to see probably doesn't help keep the signal-to-noise ratio down for them. Nor does the likelihood that a fair number of users are hitting it for everything as a protest.
- What about the "Adult Concepts" setting - I don't think I saw that in the flag-someone-else discussion? That's because it's not there. Even LJ doesn't really want to go there, so that's a strictly optional setting at the original poster's discretion rather than something they'll try to impose.
- Can't this turn into one more way for people who just don't like me to "punish" me by making spurious reports about my posts or journal? There are two factors that should make this less of a problem long-term: first, the ultimate call of whether to do anything gets made by LJ, not by the people reporting your post/journal. Second, they have a (vaguely described) mechanism for reweighting reports from people who've made spurious reports so that they might not even trigger the review queue. Other than that, they don't seem to go beyond that to any sort of punishment for trying to abuse the reporting functionality.
- Do I get notified if my post or journal is flagged? It currently looks like you don't get notified if one of your posts is reset on you after the fact, which is going to be a sticking point and I wouldn't be surprised to see them change that. It sounds like you do get notified if your journal or the community you run gets reflagged.
- What if I disagree with the judgement call? You can file an appeal through the normal abuse channels and it will get reviewed by "different eyes". Again, your optimism may vary, but that's about what you could expect unless they restructured the approach completely.
- Where's the flag-someone-else button on a friends-locked post? There isn't one - they're only letting third parties flag posts that aren't friends locked. People going all-friends-locked (and the even larger group who already lock anything that would be "TMI" or "NSFW" to start with) shouldn't see this third aspect much at all.
I mentioned "attempt to keep kids from seeing this" but they may also collapse some forms of adult content even for known adult users so that it behaves like it's behind a user cut. If you want to turn this off and always go straight to the "good stuff" (especially if you have friends who are flagging everything as adult content as a protest) you can turn this off in your viewing options.
Finally, key to all of this, there's the whole matter of how they tell how old you are. The answer is they use the birthdate you told them when you signed up, or the birthdate you told them in your profile, or the birthdate you told them the first time you tried to look at something that was automatically hidden as "may contain adult material". The common factor, of course, is "the birthdate you told them" which is something no minor ever lies about - although it does let them say "but we tried, and they said..." if challenged later. Which, for everything I point out above that makes this at least a little less screwed up than some of the louder response suggests, reduces all of this to the one question:
If this is all ultimately meaningless CYA, is this really the least complex and inflammatory implementation they could come up with?