{ thinking out loud about the things i care about }

Archive for April, 2009

Jonathan Haidt and the Moral Matrix

Sunday, April 26th, 2009

It has been a Ted Talks kind of weekend.

I particularly enjoyed this one by Jonathan Haidt, wherein he discusses the moral matrix and the real difference between conservatives and liberals. As anyone who knows me really well is aware, my scale of morality has no end caps and tends to look deceptively like globs of swirly many-toned gray paint. Obviously, this talk appeals to that intrinsic part of my nature in many ways, least of which is how my political pendulum swings depending on the topic and the weather.

Comments Off on Jonathan Haidt and the Moral Matrix

Dreamwidth and the Ad vs. Subscription Social Media Business Model

Tuesday, April 21st, 2009

There’s a hullabaloo! The hot topic in fandom circles on Livejournal right now is Dreamwidth and whether or not to move. The closed beta and impending open beta is prompting all sorts of conversations, and since I’ve opened up a Dreamwidth account I wanted to talk a little about why I’m interested in the project.

It’s not for ToS reasons. Let’s face it, no matter what is promised or who promises it, at some point all Terms of Service agreements start to look the same. As usual, the devil’s in the details. No service can legally allow anyone to publish child pornography, for example, but what qualifies as child pornography is the tricky bit. My guess is this will continue to be some fanartist’s battle for a long time no matter what service they use.

For me, Dreamwidth is all about their ad-free business model.

Not because I mind ads, necessarily, though I know a lot of people are uncomfortable with a service that makes money of their content. It’s because making money off ads on a social media site doesn’t work and is increasingly proving to be an unviable business plan.

Millions of people use social media tools like Facebook, Twitter, and Livejournal every day. Thousands of gigabytes of data is created and stored on their servers, and every day this data is repeatedly uploaded and downloaded by millions of people. Operating social media sites is not cheap, yet across the internet the service is offered for free. Why? Because, with all those hungry eyeballs, social media looks like a match made in heaven for ad serving.

Except it’s not. Ads (text or banner) sell on social media sites for dirt cheap. Why? Because their clickthrough rate is abysmally low, and in order to make the ROI (return on investment) worth it, you have to sell hundreds of thousands of impressions for a significantly reduced price to compensate for the low clickthrough. Social media sites practically give away their ad space. They have to, because otherwise no one would buy it. As a result, social media sites attract the cheapest looking ads from companies or people who have very little to spend on space. That’s why MySpace and Livejournal are chock full of brightly coloured, poorly designed, blinking scam-ads (like the plethora of YOU’RE THE ONE MILLIONTH VIEW AND YOU’VE WON AN IPOD/TV/LAPTOP ads) by companies you’ve never heard of. They’re the only ones who will buy them.

The question then is: why don’t people click on ads on social media sites? Ad revenue has become a silver bullet for search engines, why doesn’t the model translate?

A lot of reasons, but one of the big ones is that when you’re browsing your Livejournal flist or Facebook feed, what you’re not looking for is to purchase something. Not usually, anyway. Spending money is probably the farthest thing from your mind. You’re looking to find out what your friends have been up to, browse your sister’s vacation snaps, or read some fanfic. Buying something is probably the farthest thing from your mind.

Not so when people Google something. A lot of the time, you’re actually Googling for a service. A release date for a DVD. A product you’re thinking about buying. A trip you’re planning to take. That’s the prime moment for advertisers because, even if you don’t consciously realize it, you’re extremely receptive to ads when you search, especially to targeted ads. The ads you see every time you do a Google search get extremely high clickthroughs and are proven to make a lot of money for the businesses that buy them.

This brings us to the second big reason why sites like Livejournal and MySpace don’t make good ad partners: no targeting.

Targeted ads are why Google ads are so popular. When you search for a term like "hotels in Florida", you get all kinds of ads relevant to that search. All the ads you see will talk about resort packages or vacation packages for Florida destinations. This is prime ad real-estate for Expedia, BookIt, and other booking agents. They’ve all highlighted terms like "hotel" and "hotel Florida" as keywords they’d like their ad to appear on, because those are the ones that are most likely to convert to sales.

Some social media sites are able to target (Facebook being a notable example) but the reality is that most aren’t, because they don’t have any way of gathering the demographic information they need or because their search tools aren’t anywhere near robust enough to handle that kind of dynamic content targeting. And even when those social media companies partner with Google Adsense to leverage Google’s search tools, the pickings are still slim because the valuable keywords don’t deliver on Livejournal the way they do on other sites, partly because Google can’t tell the difference between the context of the term or why it’s been used. There’s a marked difference between the ROI of an Expedia ad that displays next to a blogger reviewing a hotel they went to and one that displays next to a curtain!fic where Ron and Hermione discuss their summer holiday plans.

There are lots of other reasons that ad-revenue is proving to be unsustainable for social media sites (trackability and companies not wanting to be associated with Harry/Snape erotica are just a couple of the others), but those are the big two. They’re the reason I don’t buy ad space for my company on social media sites, at any rate.

Which brings us back around to Dreamwidth.

I like their business plan. I’ve been thinking for some time now that the only way social media sites are going to survive the next bubble is if they make a tiered subscription system that is built around attracting users to pay for the service. Not everyone, because that’s a pipe dream, but a higher percentage than do today.

That’s the biggest difference between Livejournal and Dreamwidth: Dreamwidth wants paying subscribers. They’re serving their users and trying to sell us things, not trying to use us to sell things to other people, and if you think that won’t affect the way they implement and enforce their ToS, you’ve misunderstood something along the way.

Yes, paid accounts will be more expensive than they are on Livejournal, but that’s because that is the source of income for the Dreamwidth team, not merely a supplement of it. $40 a year for a premium account at Dreamwidth breaks down to just over $3 a month, and the lower $25 paid account works out to just over $2. We’re talking about the price of one Starbucks coffee (or two regularly priced coffees) a month for unlimited journal storage, unlimited bandwidth, more user icons than you’d ever really need, plus a swath of other features (some available now, some coming down the pipeline soon, and others on the roadmap), and a company that won’t side-step my interests to make themselves more attractive to advertisers. That, as far as I’m concerned, is a steal of a deal at twice the price.

Only time will tell for sure whether or not Dreamwidth’s plan is truly viable, and I hope it is. I like to support the things I use online and wish more people would remember that these sites are businesses and services. Everything you do on Livejournal — even if you’re just a lurker who doesn’t have an account — costs people money. If they can’t recoup the costs, then the service will go away. It’s really that simple.

Comments Off on Dreamwidth and the Ad vs. Subscription Social Media Business Model

Some Thoughts On Amazonfail & Some Final Stats

Tuesday, April 14th, 2009

Day One | Day Two | Day Three (Final Stats)

A recap for those following along at home:

This may not be my final post on the topic, but with so many news articles and blog posts circulating, it’s essentially impossible to keep track of them all. I may post a final recap — mostly for my own benefit as someone looking to write up a comprehensive case-study on the topic for professional reasons — later in the week, but unless something new develops, this is probably my final comment.

Obviously the first half of this story — that GLTB, feminist, and disabled sexuality books, many of them not containing any erotic content, were flagged as "adult" and stripped of their sales rank, affecting their searchability on Amazon — is well-reported and what caused the internet to rise up and strike at Amazon. And while I share that outrage, many people have spoken about why this matters very eloquently elsewhere, so forgive me if I switch to more professional gears for a moment.

The second half of this story is that news of this broke in social media, primarily Twitter, and Amazon was already attempting to fix the problem long before mainstream media news outlets even got a hold of the story. Also puzzling is why Amazon’s PR department allowed this upsurge of bad PR to rise unchecked for so long, and why they are still continuing to allow bad opinion to circulate largely unchecked after their short and altogether unsatisfying statement about the issue without getting into the game. For a company that uses social data to great effect on their site and is considered one of the great Web 2.0 pioneers, their blasé attitude toward the negative up swell is at the very least shocking, and at the most dangerously ill-advised. I suppose we’ll have to wait to see how their pre-canned comment strategy works out for them as things die down (or perhaps don’t die down) over the next couple of days.

Was it a glitch? I think that’s mostly spin. (And so does most of the rest of the internet.) At the very least I think this was probably and badly thought out attempt to "protect the children" without fully understanding their own complicated and increasingly irregular tagging and category structure. As an e-commerce professional specializing in usability, I can say in my professional opinion that it’s a good thing most (if not all) people use Amazon’s search tools to find what they’re looking for, because their catalogue hierarchy is nightmarishly inconsistent, with different editions of the same book having different tags (some examples of which have been highlighted in this excellent post on Amazon’s meta data).

This also may go to show how middle-management of a large corporation can be a flimsy creature indeed, and how someone in the middle-to-upper echelon of a large company can perhaps push through changes without considering their full implications or spending the time to do a thing properly. Was it a knee-jerk response to a right-wing trolling effort, similar to the infamous Livejournal strikethru incident? Was it a hacker? Was it a policy shift that got executed too quickly and very badly? Was it a translation user-error made by a French employee? As long as Amazon continues to be tight-lipped on the topic, we can’t know for sure. Certainly anything that comes as an official statement from them will be painted up and spun round.

I tend to think it was another one of those unintentional things that reflects underlying social privilege and inclination to misunderstand and misrepresent those things that are different from so-called mainstream. Having said that, it was uplifting to see so many ordinary people rise up so quickly and say: no, this is not acceptable, and this is why. Rest assured that Amazon felt that slap, and hopefully other big corporations (and even governments) will think twice before implementing similar changes without a significant amount of research, thought, and openness.

Good job, internets. Sometimes you use you’re flaily powers for wank and eye-roll worthy silliness, but this time you have used your powers for good. Well done.

ETA: I’m going to link some "fallout" articles here, mostly for easy finding.

ETA: I’m going to link some "fallout" articles here, mostly for easy finding.

Comments Off on Some Thoughts On Amazonfail & Some Final Stats

Amazonfail: Day Two

Monday, April 13th, 2009

Day One | Day Two | Day Three (Final Stats)

Amazon still insisting it was a glitch but hasn’t released any sort of detailed statement, just “we’re working on it”. Not exactly what the internets want to hear right now. The amazonfail hashtag search is just as hopping as it was last night. As of this post, wthashtag is showing almost 130,000 twitters tagged with amazonfail (over 30,000 already this morning), the petition goal post has been moved again and lists over 13,000 names, and more of the media has picked up the story including CBC, The Guardian, Entertainment Weekly, The Wall Street Journal, boingboing, The New York Times, The Advocate, and The Huffington Post.

Gawker wants you to know why they don’t buy the glitch line, and neither does most of the internet. Queerty is not impressed by excuses. Change.org has started a petition-like letter-writing campaign over on their site as well.

For people who don’t get why the internet exploded. After Ellen has a even better article about why this matters, the over-sexing of GLTB people, and what the larger problem is that everyone should read.

There’s a good round-up by right here on Livejournal. Also one from . And has an excellent, well-written summary. Some thoughtful thoughts from . And another summary from .

Dear Author looked at the metadata of each book and discovered the probable data links are exactly what the internet thought:

I looked up over 40 books that had been deranked and filtered out of search engines.  It appears that all the content that was filtered out had either “gay”,  ”lesbian”,  ”transgender”, “erotic”  or “sex” metadata categories.  Playboy Centerfold books were categorized as “nude” and “erotic photography”, both categories that apparently weren’t included in the filter.  According to one source, the category metadata is filled in part by the publisher and in part by Amazon.

Top Google news result for search term “amazon” gives you amazonfail topics by major news outlets. Unsurprisingly, searching for “amazonfail” gets even more hits, though last night both search terms only elicited a handful of top results on the issue. The Amazon Rank Googlebomb is now the top search result for term other than the news about amazonfail. A blog search shows over 6,000 results for the term amazonfail.

Over 1,000 affected books have been tagged on Amazon with amazonfail (at least it will be easy for Amazon to find the affected books).

The merchandise has landed. So has the lolcat.

Social media experts are starting to take note and document the amazonfail case study.

And if you think this is a quick fix and are wondering why everything’s still de-ranked, explains why it isn’t.

ETA: Salon article. And, from their excellent followup article:

At the very least, the “glitch” line suggests that this wasn’t supposed to happen, and Amazon recognizes it’s a highly undesirable situation for the company. Whether that means, “We had no intention of discriminating against anyone” or “We had no intention of so many people figuring this out at once and dragging our brand name through the mud” is an open question.

Also, ITWorld wants to know, and rightly so:

Getting back to my point, it’s dumbfounding that Amazon would let this controversy grow unchecked for a whole weekend. For such a giant in the online space, they certainly seem to be behaving like a brick and mortar company from two decades ago. The fact that this past weekend was a holiday for many people doesn’t stop the flow of information on the internet. Someone at Amazon HQ is going to have a very bad Monday, and deservedly so.

ETA2: Affected author Heather Corinna blogs about amazonfail on her Amazon blog. Meanwhile, wonders about the pre-canned nature of customer service responses that may be complicating the way Amazon responds.

ETA3: This guy says he’s responsible. After some clever detective work, people think not so much and call troll.

ETA4: Apparently people will try to make money off of anything.

And Reunifygally wants to remind us it’s not just GLTB books that are affected, but also books about sexuality and disabilities. I recently saw a performance of the Vagina Monologues that added a section on the disabled and how they are “protected” by their guardians and caretakers from their sexuality. It profoundly disturbed me.

ETA5: Neil Gaiman posts on his blog about amazonfail.

ETA6: Getting reports that amazonfail is breaking on Channel 4 TV news. I don’t have TV, but it’s all over twitter. (ETA: Report has been posted online here. Fast-forward to the 8 minute mark to see the amazonfail report.) It’s also starting to gain momentum on digg. There is an amazonfail tag on social bookmarking site delicious. And from Amazon’s own twitter feed? Business as usual.

ETA7: Affected author Jessica Valenti calls her editor who contacts her Amazon rep, who notes that this is no glitch.

ETA8: GLTB books seem to be re-appearing in the bestseller listings as of around 1:15pm MST today. And more conversation about the PR disaster side of the story.

ETA9: Reports of an anonymous Amazon coder insisting it was a “real person” who mass-changed the tags of over 58,000 books, though anon. qualifies that they can’t verify if it was intentional or accidental. Other reports indicate it may have been some sort of wacky translation error or an employee user error.

ETA10: Amazon spokesperson Drew Herdener admits error that affected over 57,000 titles in several categories:

This is an embarrassing and ham-fisted cataloging error for a company that prides itself on offering complete selection.

It has been misreported that the issue was limited to Gay & Lesbian themed titles – in fact, it impacted 57,310 books in a number of broad categories such as Health, Mind & Body, Reproductive & Sexual Medicine, and Erotica. This problem impacted books not just in the United States but globally. It affected not just sales rank but also had the effect of removing the books from Amazon’s main product search.

Many books have now been fixed and we’re in the process of fixing the remainder as quickly as possible, and we intend to implement new measures to make this kind of accident less likely to occur in the future.

Also, someone screencapped this on the Twitter homepage. LOL. And new definition for Amazon Rank has made it into the Urban Dictionary.

More thoughts from The LA Times here. Amazon Spokesperson Drew Herdener’s quote has also been picked up by The Seattle Times and The LA Times.

ETA11: Richard Eoin Nash on the social contract Amazon violated during amazonfail.

Comments Off on Amazonfail: Day Two

Amazon and Homophobia: Day One

Sunday, April 12th, 2009

Day One | Day Two | Day Three (Final Stats)

Amazon is suppressing sales rank information, affecting the searchability of LGTB books, calling them "adult". For a list of books that have been affected (includes quite a few young adult books and biographies as well as some critically acclaimed fiction) go here. For a growing list of twitters on the subject, peppered with links throughout, try #amazonfail. Here’s a thoughtful link explaining why this matters. Someone has made an AmazonFails logo. And a petition, if you sign those. It’s always better to take the time and mail a letter direct, though:

If you want to protest to Amazon direct, this is their email addy: connect-help@amazon.com
Other ways to get in touch with them: http://clicheideas.com/amazon.htm

Also try ecr@amazon.com and the customer service phone number is 1-800-201-7575.

Or write to their CEO:
Jeffrey Bezos. (total annual compensation $81k – irrelevant but interesting). The address is 1200 12th Avenue South, Seattle, Washington 98144-2734, United States Phone: 206-266-1000 Fax: 206-622-2405

ETA: Google Bomb FTW! New definition of Amazon Rank. I believe there is a movement underway to get this definition into the Urban Dictionary. Also also: Here’s a good base letter to start you off.

ETA2: Apparently Amazon Customer Support is being completely overwhelmed by this. Happy Easter Sunday, Amazon! People scoff at the power of social media. At some point in the next 12 hours as this hits the mainstream, the second half of the story will be how this broke on Twitter and no one else was there for hours.

ETA3: Apparently even children’s books like Heather Has Two Mommies have been stripped of their rank. Amazon, meet PR disaster.

ETA4: Starting to break into mainstream media. LA Times Blogger has picked up the story. Also, it’s appearing in Amazon’s own forum threads.

ETA5: As of 4:30pm MST, the amazonfail hashtag Twitter search is updating way too fast to keep track of. Someone at Amazon is going to have a really, REALLY bad day tomorrow.

ETA6: Amazon cries glitch. (Traffic has killed the site; for screencap, go here.) But other people say this has been happening since February. Who knows at this point. Elsewhere, has a possible theory that is not entirely out of the realm of possibility, but we’ll see what Amazon’s PR department comes out with. In the meantime, the amazonfail hashtag is updating at the rate of about 75 tweets per minute or so (estimated by me as I watch the number slide up in the tab as I make this edit). Also, popular site After Ellen has caught the scent.

ETA7: Amazon continues to be cagey about the "glitch", even with the LA Times. They really need to get out an actual statement about this, sooner rather than later. Likely that will get picked apart, too, but at some point soon they need to start playing defense instead of this incessant no-commenting. This is the kind of thing that really needs a comment. Users are now tagging as many of the affected books on Amazon with the tag amazonfail.

In addition, for anyone who’s interested, I started tracking amazonfail at around 11:00am MST Sunday morning. At that time, a wthashtag.com search showed around 1500 tweets with the hashtag #amazonfail. It’s now 10:30pm MST Sunday night, and that same search shows over 80,000 tweets. When I checked the petition at around noon it was at just over 50 signatures. The same petition just passed the 7,600 signature mark.

ETA8: The Associated Press has also picked it up. So has The National Post. And CNet News. Now we’ve really got us a party!

ETA9: Interesting post about the real dangers of this recent set of events combined with an entirely digital world via the Kindle.

Jezebel has a really great write-up with a good time line tracking the events. And with that, it’s bed time for me.

Comments Off on Amazon and Homophobia: Day One