Commerce

To your point, Loving - and I suggested this in the Commerce Group Chat -- it would be far more effective to employ a filter that acts more like a spam rating filter, responding to multiple occurrances of words flagged in the Adult and Moderate blacklists. For instance, a single instance of the word "cock" or "slave" (both of which have perfectly valid, non-sexual definitions in the dictionary) should not result in a Moderate or Adult rating. However, some number of flagged words would result in the listing receiving a more mature "ranking;" for instance, a ranking of 3-5 might result in a Moderate rating, a ranking of 6-8 Adult, 9-10 might be banned content (think "lolita" "child" and "sex" in the same listing). Certain words might trigger a Moderate or Adult rating all by themselves, but that should be reserved for words which have only one potential interpretation or definition.

Anyway, just an example there. The points are:

1. Except in rare instances, a single word in and of itself should not trigger an Adult or Moderate rating -- and definitely not when the word has perfectly valid non-sexual definitions. If you want to use a computer program to make those kinds of judgements, some intelligence MUST be built into it, otherwise ridiculousnous prevails...and ridiculousness is not conducive to business.

2. Personal responsibility for our own content MUST be part of the equation. If someone with Moderate content -- or more importantly Adult content -- is purposely trying to avoid an accurate rating that reflects that content's nature, then punish the offenders, rather than making 100,000 users jump through hoops to comply with a system that's vague, inaccurate, and open to various interpretations across the globe.

3. Publish the flagged word list. It is ridiculous in the extreme to make us waste our time guessing what might be in that list, for reasons I've stated previously.

Rachel Darling wrote:

1. Except in rare instances, a single word in and of itself should not trigger an Adult or Moderate rating -- and definitely not when the word has perfectly valid non-sexual definitions. If you want to use a computer program to make those kinds of judgements, some intelligence MUST be built into it, otherwise ridiculousnous prevails...and ridiculousness is not conducive to business.

Indeed, some words have mutiple meanings, context is important, Pamela's bird that cocks its head is a perfect example of why just focussing on a word is bad.

Rachel Darling wrote:

2. Personal responsibility for our own content MUST be part of the equation. If someone with Moderate content -- or more importantly Adult content -- is purposely trying to avoid an accurate rating that reflects that content's nature, then punish the offenders, rather than making 100,000 users jump through hoops to comply with a system that's vague, inaccurate, and open to various interpretations across the globe.

 

Absolutely, this can't be said often enough. I would hope that we all want an accurate rating system, there are of course grey areas but those can be ironed out.

Rachel Darling wrote:

3. Publish the flagged word list. It is ridiculous in the extreme to make us waste our time guessing what might be in that list, for reasons I've stated previously.

Well if they won't publish it, at least highlight which word is causing a problem in that particular listing, people shouldn't have to work out which seemingly inoccuous word is the problem. They say they won't publish this list because people will just get around it, but that takes us back to point two and those who are actively trying to get around the filters, will work it out for themselves anyway, as they know they're trying to circumvent the filters whilst those who are genuinely puzzled why their PG item isn't getting listed, are the ones being punished.

Rachel Darling wrote:

1. Except in rare instances, a single word in and of itself should not trigger an Adult or Moderate rating -- and definitely not when the word has perfectly valid non-sexual definitions. If you want to use a computer program to make those kinds of judgements, some intelligence MUST be built into it, otherwise ridiculousnous prevails...and ridiculousness is not conducive to business.

http://blogs.secondlife.com/message/592050#592050

Bingo.

As a seller of 'Adult Products' I have no wish whatsoever to corrupt the under 18 market by exposing them to my products.

But I do want adults to be able to browse my items without having to hunt down the tiny 'enable adult content' button. The number of times I have logged on and forgotten to press this miniscule button far outweighs the times that I automatically do, and I think that I'm not alone in this. Since the new rules came in place my sales have nose dived...

When logging on there should be an entry screen which clearly asks the viewer to state adult or under 18 in line with all sites that carry adult content. By squirelling away the adult content of the Marketplace it totally skews the audience to the teenage sector by default. This is why merchants have always tried to 'cheat the system' If your products are in the general catagory, you know all your adult market are guaranteed to see them.

Let's not be coy about this...teenagers are going to press that over 18 button in any event. They always have done, and always will do. There's nothing anyone can do to stop this. But by hiding thie adult section away, and making the grown up population of sl feel like second class citizens, it damages more than it cures anything...

Nothing to say!

Bump!

We need a list of words that are no-no's. I am tired of guessing which words I can use and which I can't. I have had several listings get switched and it's a big clusterbang trying to figure out what's ok and what's not!

Why isn't it a one time thing this enable adult content .. It seems strange to block half the content of the market unless checked .. and honestly.. how does this stop kids seeing horrifying pictures of medieval houses ..As I've always said .. Adult rating are pointless unless you can stop people shouting obscenities in PG zones ...*sigh *  ( anyone noticed you're more likley to hear the 'c' word in a pg area than an adult one lol ..,, funny ol' world

Here's an update. Pasting my previous action items with details

On the Linden side, here is what we’re doing:

• Currently working on a bug fix that incorrectly flags some items as moderate and adult. We plan to release this fix on Tuesday.

This will be going out tomorrow am. Once it is released, I'll drop a note in here. We'll be asking a subset of merchants to retry saving their listings to see if this helps.

• Reading through the forum posts and filtering out comments that appear to be issues.

I was caught up as of this am. I will do my best to skim over additional comments, but, mostly, I am focused on getting any fixes identified and rolled out.

• Reviewing   the decisions, processes, testing, etc. done prior to the roll out to  identify actionable changes that can be made going forward.

We had an internal post mortem today to identify what we can do better as a group in the company to make things go more smoothly for us and for the Merchants and Residents who use Marketplace. I'll be presenting a summary during the Marketplace User Group on Thursday am.

• Monitoring   gross market value for the Marketplace as a whole and the listings  that  were migrated to determine if we are seeing a change in behavior.  At  this point we do not have enough data to make an assessment and  will  review on Monday.

We did see a dip on the items that were moved, so we are changing the default when you are over 18 and logged in to Marketplace to show you Moderate content in addition to General. We are doing our best to get this out before the weekend. I had a very last minute conversation with some merchants on this to make sure this made sense. I've posted the chat here. The people in this chat were pulled based upon certain data points and some last minute grabbing of people online and from a previously existing user group. People giving feedback will be rotated--so you'll get your chance!

• Reviewing listings that you all have sent to us and modifying our filtering to fix any issues.

We have been reviewing CS tickets and the JIRA items. A subset have been passed on to dev for further investigation. We are going to keep the same JIRA open, per the request of CS. So please continue to post there.

• Working on getting phase 2 rolled out (advanced search options) to make it easier to find listings marked as moderate and adult.

I'll be setting up a review later this week or early next week to look at what we are proposing. This will be with a targeted user group of merchants. I'll also take a subset of people who do not fit the data criteria but want to participate. I'll discuss the criteria we are using during the Marketplace User Group on Thursday and provide a means for signing up.

On the email issue: if you asked Marketing to remove you from their lists, you will not get the email that was sent. That is the means I currently have available for sending out mass emails, so opting in will make sure you get these communications from Marketplace. Yes, there are other solutions we could put into place to send mass emails, but the same information will be posted in this forum and tweeted. Since there are other options, I will prioritize this below software fixes for now.

Brooke

Brooke,

Moviing Moderate items to General is a start - helps my items that were flagged to moderate even though they are completely G Rated.

Hopefully this "bug fix" fixes things.... but i cant see how if you are using the basis of flagging/categorizing items by their KEYWORDS... its just a wrong strategy to start with.  That is the bug -  that you are using keywords.

As for email issue... I and many others asked to be taken off LL's Spam Marketing email list when in 2009 they were using our Customer Email Contact info to promote 3rd party merchant and vendors to promote their events.  This should not have taken us off legit LL important customer email communications.  And... it doesnt seem it has in some cases.  Your team sent out an email to merchants in early January and I received that email - so why not this one?  Are you switching between lists?  Your customers should not have to endure LL Markeing email spam in order to get serious important customer service email communications.

I am glad you are trying to address these issue and I really do hope you follow up with action on addressing this terrible deployment.  Here is hoping your team recover the damage from this one.  Sadly I dont think LL will allow your team to stop this Child Protection Filter strategy.  Its wrong and LL will lose the larger population of their customers to protect a very small fraction of teens who should not have been allowed to enter.

Thats another story and before your time.

Good Luck.  Wish I could attend your open office but not if you will be having them all in mid day North American time.

I am just curious how they are to determine ratings without using keyword filters at all.

Brooke Linden wrote:

On the email issue: if you asked Marketing to remove you from their lists, you will not get the email that was sent. That is the means I currently have available for sending out mass emails, so opting in will make sure you get these communications from Marketplace. Yes, there are other solutions we could put into place to send mass emails, but the same information will be posted in this forum and tweeted. Since there are other options, I will prioritize this below software fixes for now.

Brooke

Brooke,

There is a problem with emails. When bulk emails are sent out not everyone gets it. And it isn't always the same people who miss out. Sometimes my partner gets an email and I don't, and then vice versa. This time I got the email on maturity and he didn't. Unless this is fixed don't rely on emails only to communicate important information. This problem has nothing to do with opting to be removed from a list. There is something broken when sending out mass emails, which has been broken for a long time, and it needs to be looked at and fixed.

 

or Twitter, a very small number of merchants use that.

Brooke, why don't you all do us a favor, and admit Marketplace is borked beyond repair and start over?

I went from having 2-3 sales a week on xstreet to 2-3 a month in parketplace when it first opened to none now because my products (all listed under an alt) are ALL incorrectly flagged because of your hidden crazy filtering scheme.  Transaction id numbers still dont match, nor does the idiotic half measure you attemped to put in in its stead.

You want to help merchants, then why not join one of the groups in world and listen to what merchants have to say? or better yet join in the push ive been making to get a dedicated team to address the needs of all content creators for the web, server, and viewer portions of SL? (look up the group sl-ccnaf).

one more thing. why was this posted to the forums instead of the blogs where you know... people would have seen it?

Just to throw in a different view: I couldn't care less if my "marketplace stalls", that I consider to look good in a Gorean sim, will be flagged mature or even adult. As long as it is remembered what people's preferences are when searching.

Every sane person immediately enables the "Yes I really REALLY (i say this for the thousand and first time now)... I REALLY want to search adult items too!!!!" Why? Simply because people do not want to be restricted in general and they will also realize soon enough that LL has made a mess from search once more and that checking it will be the only way to be able to make them find what they want. Why would I care if LL looks retarded with flagging my items as such?

As long as people can find it! So... it should be remembered what preferences someone has
<sarcasm>oiiii.. rocket science here as that would perhaps require to send out cookies... must be difficult for LL</sarcasm>
As long as it is remembered then 99% will all check it and the teens can not check it. Sure... let them look for Gor... it won't show up. I couldn't care less if I lost those 250 teens as customers. I didn't want them here in the first place and they don't want to spend money anyway. So... please please please make it so that accounts remember preferences and I don't give a damn about the silly things you throw at us here this time.

If you read the chat transcript, this is what Brooke is planning to do.

Cummere Mayo wrote:

 

one more thing. why was this posted to the forums instead of the blogs where you know... people would have seen it?

 

This is a blog, btw.

Brooke asked for feedback on how to best communicate with merchants. Not many said "The Blog". Did you?

You have a point. The problem is however that you have to be logged in to view the 'adult' search. I do not have any numbers on how many guest/not logged in people are on the marketplace, but on SLX is used to be between 500 to 1500 constantly, thats a lot of people that wont see your mature/adult listings...and wont be able to click on your inworld location.

Bootsy, read the chat Brooke posted. Prefs will be persistant. Problem solved.

How can a simple (and totally PG) hallowen decoration table be forced Adult?

 

Jeez, please test such systems before forcing them upon users.

Jaylin Wytchwood wrote:

 

 

Jeez, please test such systems before forcing them upon users.

 

Did you read the part of Brooke's message where she addresses this issue?

You know, those of us who are here on a daily basis appreciate the vastly greater level of communication that Brooke is attempting. Those who drop in once in a while may not be so aware of the dramatic change.

Thanks for the candid feedback Brooke it's much appreciated. Persistant logins is the way to go linked to our inworld prefs ideally. Given that there  is a considerable problem  concerning many crossover items being flagged as more mature than they should be and a bug that makes  some clearly adult thingys PG perhaps it may be best to call a moratorium on the start of the Flagging Festival which was due to begin today.

^L^

Pamela Galli wrote:

 

Bootsy, read the chat Brooke posted. Prefs will be persistant. Problem solved.

Read it just now. Well it does give hope, especially the part that mature is not treated the same as adult anymore. I still have my doubt about the word filter, I think it will remain a pain in you know what. For every new listing you create you will have to save the unfinished listing after every sentence to check if there isn't any word triggering the filter.

Writing down a 5k description and then saving it, finding out that there is a faulty word, having to check all words, re-save the listing a zillion times trying to find the word before it is actually rated as G will result in whole bunch of swearing behind the PC.

I am glad that there is a jira now and a open communication. We will see what happens this week.

Thanks for the update. On the subject of emails, I for one have NOT opted out of anything and still have NOT received the email regarding this maturity change. So, it's not just those who have opted out...it seems to be random as to who receives or doesn't receive emails.

I would like to also add..this is not only to Brooke but to the commerce team, and especially those who oversee them...

Please don't mistake the results of your actions as general merchant hatred/distaste for LL in general.  It's just that we have been shouting issues at LL for months now.  Since MP went "live" (I use the term loosely as it still operates as a beta product IMO) we have filed jira after jira.  We have begged and pleaded with any Linden that would listen to fix known issues that are plaguing the merchant community.  We have begged and pleaded for known issues to be examined and fixed.

Yet the first public act of the "new" commerce team has just added to our issues.  We are fed up, frustrated, disillusioned with the whole thing.  Had you fixed MP search, come up with a short term fix for delivery failures, made adding listings easier, addressed any one of the numerous other issues/complaints about MP (publicly) before swinging your mighty filter around, we might be a little more inclined to say "ok, this was a MAJOR screw up but at least they fixed XYZ."  Instead, your first public act was to reign down a mess eqaul to the initial migration from SLX.

And "sneakily" too, I might add.  I, for one, do not recall ANY mention of a filter being used when the new system went into effect.  You were clear that anything marked "mature" would be moved to adult.  That didn't seem unreasonable to us and we all were aware we might have to go in and fix anything that was mature before and decide if it was moderate or adult.  We were given NO indication there would be a filter put into place.  Nor was it indicated that the judgements of this filter would be irreversible without the hoops we have had to jump through.  Bad form.  And shame on you all!

It's as if we are kids who have been left home alone countless times then all of a sudden, mom and dad have decided we need a babysitter.  We didn't do anything wrong.  We've done well at keeping the house clean, didn't throw parties, didn't eat everything in the fridge.  But now we have a babysitter telling us what to watch on TV and when to go to bed... it's insulting, quite frankly.  The merchants have always had the ability to police their own items.  And for those who tried to "cheat" the system, we've been able to flag them and have action taken.  Peers policing peers has worked up until now.  I could understand if this came after a public request from LL to properly rate our items "or else".  But there was no indication in advance this would happen.  And it's just a shame that this is our first experience of the "new" MP team.