The Long Tail

Can you make money in the Long Tail?

Well, it depends who "you" are. If you're an aggregator, sure. But if you're like the majority of Long Tail microproducers, direct revenues can be harder to come by. I was reminded by this Valleywag post that one of the most common misunderstandings of the phenomena is that it somehow makes it easy for individuals to translate low popularity into riches. Here's the key part of the post:

I'm beginning to have my doubts about Chris Anderson's long tail, the proposition that cultural boutiques can make a living on the Internet. One disgruntled publisher complains she's owed less than the minimum Google can be bothered to pay her. And, as fast as she makes money, Google lifts the threshold. [She writes:] "When I started with Adsense in late 2004/ early 2005 the minimum was $25. Just when was about to hit the $25 minimum, they raised it to $50. Now that I have $45 in my account, the minimum is $100. Granted, I have a site with very low traffic, but how many website owners are getting screwed by Google? If the long-tail theory holds out, there could be millions of dollars of unpaid Google ads."

First, let's review what the theory actually says. The nutshell version of it is this two-parter: A) if you can dramatically lower the cost of production and distribution, you can offer far more variety. B) given more variety and the tools to easily organize it for individual taste, people will increasingly revel in their differences, rather than settling for their commonalities as in traditional blockbuster culture.

In Long Tail markets, hits lose their monopoly on culture as they share the stage with million of niche products. Minority taste rules.

There are three basic types of participants in Long Tail markets: consumers, aggregators and producers (note that it's possible to be all three; these aren't mutually incompatible). The main effects on each are:

• Consumers. Effect: Largely cultural. People have more choice, so individual taste increasingly satisfied even if the effect is an increasingly fragmented culture. 
• Aggregators. Effect: Largely economic. It's never been easier to assemble vast variety and create tools for organizing it, from search to recommendations. Increased variety plus increased demand for variety equals opportunity. Also note that just as one size doesn't fit all for products, nor does it for aggregators. I think the winner-take-all examples of eBay, Amazon, iTunes and Google are a first-inning phenomena. Specialized niche aggregators (think: vertical search, such as the real estate service Zillow) are on the rise.
• Producers. Effect: Largely non-economic. I responded to a good Nick Carr post on this last year with the following:  "For producers, Long Tail benefits are not primarily about direct revenues. Sure, Google Adsense on the average blog will generate risible returns, and the average band on MySpace probably won't sell enough CDs to pay back their recording costs, much less quit their day jobs. But the ability to unitize such microcelebrity can be significant elsewhere. A blog is a great personal branding vehicle, leading to anything from job offers to consulting gigs. And most band's MySpace pages are intended to bring fans to live shows, which are the market most bands care most about. When you look at the non-monetary economy of reputation, the Long Tail looks a lot more inviting for its inhabitants."

So to answer the publisher quoted by Valleywag, the Long Tail never promised you Adsense riches. If what you're doing has value, it does promise you more attention, reputation and readership. But converting that non-monetary currency to actual money is up to you, and there are as many ways to do that (from better job offers to consulting) as there are people who wish to try.

As for Google sitting on the float on millions of dollars of sub-threshold ad payments, well, Google is the canonical Long Tail aggregator. By capping minimum payments it's clearly found a way to improve the economics of serving the Long Tail of advertiser, advertisement and ad-driven publisher. If that's not what such Long Tail publishers want, that suggests that there's a nice opportunity for a competing ad network that can serve them better. One size doesn't fit all.

Speaking of radical transparency...

I read first on Reddit co-founder Aaron Swartz's blog that he and Quinn Norton are going to be geo-trackable in realtime as they cover a hacker convention:

"This year's Chaos Communication Congress opens with a unique opportunity -- your chance to track the movements of a Wired News' reporter on the scene, as well as nearly a thousand other visitors to the annual hacker convention.

[...]

track Quinn Norton (ID 254135) and Wired Digital's Aaron Swartz (ID 254260), two of nearly 1,000 participants wearing RFID trackers for the duration of the congress.

(Quinn Norton, Hacker Con Submits to Spychips, Wired News, December 28, 2006)

Swartz, who is now a Wired employee thanks to our acquisition of Reddit, adds: "It's funny--the first time my employers find out where I am this week may be when they report it on their own news site." Which may in fact be true--like many email/IM based companies these days, we increasingly don't care where people work as long as they're Treo/Blackberry reachable and responsive. Or just update their blog frequently ;-) 

Our competitors are invited to spend a lot of time writing geotracking scripts to see who our two intrepid reporters are interviewing.

Briefly noted from Australia

I've been near Sydney all week visiting family, and also testing my theory that little kids adjust to time zones at the rate of one hour per day. Sure enough: they woke up 3:30 am the first day, then 4:30 am, then 5:30 and then....damn. The sun rises here at around 5:30am, the weird birds start mocking us with monkey noises and the kids are up and ready to party. There's nothing we can do to get them to sleep any longer. So much for the whole R&R thing.

I should be updating my sidebar, but I've been so spoiled by the Microsoft Live Writer, which doesn't support Typepad's Typelists, that I won't edit in a browser anymore. So here, in post form, are the top five interesting things I've just read as I caught up with my feeds.

1. Nick Carr has another astute post that analyzes a recent study on concentration in web traffic that seemed to contradict the Long Tail trend of decreasing concentration of the hits. Richard MacManus found that top ten sites accounted for 40% of total Internet page views in November 2006, up from 31% in November 2001. But Nick rightly noted that two of the biggest of the top ten sites--MySpace and Facebook--are actually Long Tail aggregators. All those individual pages on the two social networks are, in a sense, just personal blogs. Take the two out, and the top ten have about the same traffic as they did five years ago. Why not less, as the theory would suggest? Perhaps because it's harder to count traffic to all those Long Tail sites that are scattered among millions of stand-alone domains. An interesting comparison data point would be the share of incoming links that go to the top ten sites and how that's changing, something I'll bet Technorati could answer. Dave?
2. Roger Ebert, in his introduction to his 2007 Video Yearbook, discusses the Long Tail of movies, and offers some of his own data: "There may be only one person in a city who wants to see a film or read a review, but because of Netflix and the Internet, that person can do it. We observe that pattern at www.rogerebert.com, where there are more than ten thousand reviews and our Web traffic statistics show that even the most popular film represents less than 1 percent of our business. As of June 15, 2006, "The Da Vinci Code" and "Brokeback Mountain" were tied at 0.8 percent of our page views; the next most requested reviews in 2006 have been for "V for Vendetta" (0.7), "X-Men: The Last Stand" (0.6), and "An Inconvenient Truth" (0.5). The lesson: People are curious about a lot of different movies."
3. Liked the book's Lego example of the Long Tail of toys? You'll love Vicale's Long Tail of action figures, a way to mass-customize characters with outfits and accessories, all with a nifty flash interface (see above).  (Thanks to BoingBoing for the link)
4. The Long Tail of Holiday Music. Says Digital Audio Insider: "Over at eMusic, of the 1,226 holiday albums in the catalog, 1,128 have been downloaded over the past month. That's a whopping 92% of the catalog!"
5. Finally, for people who can't get enough Netflix data, Ilya Grigorik has done some useful analysis of the huge ratings database Netflix released as part of its recommendations contest. Meanwhile, Anita Elberse from Harvard Business School has been doing the best Netflix data crunching around (she and I have proprietary data from Netflix, which gives us some unique insights that you can't get from the public data), and you can read an interview with her about this research here. Look for more from her and me on this early next year, as we now have new data sets from both Netflix and Rhapsody that will allow us to take Long Tail analysis to the individual subscriber level, something that I don't think has ever been done before.

"Hardware wants to be software and software wants to be free"

Speaking of the the Economics of Abundance...Nick Carr coins another great phrase in a post about IBM giving away a software competitor to the Google Appliance.

The Long Tail of, er, Photofinishing

There are no doubt plenty of industries where Long Tail effects don't play out in any real way, at least at the product level *. Consumer packaged goods, for instance, can only be made in so many variations and Detroit carmakers, who abhor unsold inventory and have never figured out how to do BMW-style mass customization, still long for the day when you could have any color as long as it was black.

But I continue to be amazed by how many non-obvious industries do turn out to have strong Long Tail potential. And the most delightful part of this is that it tends to be the people in those industries that do the analysis and bring it to my attention, fully formed. One of the best recent examples of this was, to my total astonishment, photo finishing.

Steve Giordano, the CEO of Lucidiom, stopped by the office last week to explain why. Lucidiom is the largest manufacturer and operator of photo finishing kiosks of the sort that you find in many drug stores and supermarkets--insert your camera's memory card or a CD-ROM and it lets you select the prints you want, sending them to an adjacent high speed photo printer. Because all these machines are all connected via Lucidiom's network, Giordano has an enviable statistical view on an industry that prints more than 2 billion pictures each year.

The traditional photofinishing business is highly hit-driven in terms of its products. The 4x6 print accounts for 91% of the photos printed in mass-market retailers such as drug stores. Even in independent photofinishers, such as Ritz Camera, the 4x6 print accounts for 77% of the volume.

But the revenues picture is very different. Photofinishers can charge a lot more for other sizes and photo packages, from sheets of various size prints to photo albums. Although just 23% of the prints at independent photofinishers are for print styles beyond the 4x6, in aggregate those niche products make up a full 52% of the revenues. Today there are more than 200 products available through a typical Lucidiom kiosk, ranging from different sizes and combinations of prints available in the store to entire birthday party kits that include matching paper plates and other table decorations that you order along with the photos-based invitations from the kiosk for later delivery to your door.

Because these kiosks are simply virtual storefronts for what is essentially a digital good (different variations of bits sent to the same printer), they've got infinite capacity for product variety. And as if often the case, the niche goods have the highest margins--it costs little more to print a photo "storybook" than a standard print, but you can charge just $0.19 for one and several dollars for the other. Margins for a 4x6 print, which sell for about $0.19 each, are just 10-15%. But the average margin for "beyond 4x6" products is a whopping 85%.

As more and more customers use these kiosks, companies such as Lucidiom are finding inventive ways to drive demand down the tail of photo products, where the real money is. Today, says Giordano, the market share of the bestselling 4x6s prints is falling by 3% every six months as customers shift to these new alternatives. A classic Long Tail market in action! Who knew?

A final interesting factoid: I had assumed that the vast majority of people printed their digital photos at home or used online services such as Shutterfly. It turns out that of the 20% of the digital photos that are actually printed--as opposed to just sitting on your hard drive--about half are printed at home. Another 5% are printed using online services. A full 45% are printed via kiosks. Surprisingly, about half of those come in on CD-ROMs rather than memory cards--people find it easier to burn a CD with photos they want printed than to upload them or take the card to the drugstore every time it gets full.

----

 (* I would, however, argue that every industry needs to have a strategy for Long Tail marketing, reaching niche customer segments through increasingly word of mouth means. But I'll save that for another post...)

Transparency reaction

Lots of reaction, both pro and con, to my riff yesterday on how radical transparency could apply to media. As is always the case, it pointed out some places where I should have been more clear or nuanced, as well as some places where I'm probably just wrong. Here are some quick expansions/clarifications, which should have been self-evident but clearly weren't:

1. This obviously couldn't apply to most of the stories. Open research and editing only makes sense for a certain kind of article, one that is highly thesis-led and involves a lot of research and collection of examples. The vast majority of Wired stories aren't that--they're either too short to make an open process worthwhile, too "voice-driven" to risk the cost to stylish writing of too many contributors, or are based on the narrative form, which also works best with a singular vision.
2. Although open research can generate useful reader input at any stage, experience shows that the writing itself is best done as a solo effort and open editing only really works pretty late in the game, after most of the heavy lifting (from architecture and scope to clarity and style) has already been done by a professional editor. We've found that open editing can improve a good piece by catching small errors or suggesting additions and example that make the case even better, but it can't save a piece that needs a lot of work.  
3. These techniques works better for relatively short-form web media than they do for the print magazine. That's mostly because magazine features tend to be long (2,500-10,000 words)  and wrapped up in a package heavy with art and design. Long-form journalism is more about narrative and story arc, which is really best done as a close collaboration between a writer, editor and designer.

Note that in my own case, although I shared most of the book research and many post-length draft passages on this blog, almost all of the actual writing and editing was done as God intended, head down at Starbucks with headphones on. For writing and editing long-form prose, focus is everything.

UPDATE: Jeff Jarvis has some smart amendments and additions to this here.

In praise of "Out There" people

A consultancy called the Attention Company has just complete a fascinating survey of what it calls "Out There" people---think Mark Cuban and other promoters of radical transparency.

They define "Out There" people as those who are more likely to:

• Value fame as an “asset”
• Willing to share certain types of
  sensitive information on the web
• Believe it is appropriate to criticize their
  organizations on the web

There’s more of these than you might think – 69% of the population agrees with at least one of these statements!

The presentation and survey results are well worth reading. Here's one sample:

Thanks to Jason Calacanis, another example of this phenomenon, for the pointer. He adds:

"Frankly, I couldn't run a business any other way. The only way to really get things done is to be out there. There is so much noise in the space, so little attention, that your best bet for building great products is being totally honest about what you're doing and letting folks help you."

What would radical transparency mean for Wired? (Part 2)

In the previous post, I described how the media landscape is changing, starting with reader expectations. In this post I'll describe what that might imply for a media organization like ours. If the key word is "participation", how could we encourage that to the fullest? If trust comes come from transparency, how might we open the entire process? What does open source media really mean?

Although I'm not promising we'll do all or even most of these things, here are some first thoughts on what a truly transparent media organization would do. (Some of these are based on my experience in open-sourcing my book research on this site, which worked great. Why not apply the same lessons to a magazine?)

 

Six tactics of transparent media

1) Show who we are. All staff edit their own personal "about" pages, giving bios, contact details and job functions. Encourage anyone who wants to blog to do so. Have a masthead that actually means something to people who aren't on it. While we're at it, how about a real org chart, revealing the second dimension that's purposely obscured in the linear ranking on a traditional masthead?

Upside: Readers know who to contact. The organization is revealed as a collection of diverse individuals, not just a brand, an editor and some writers.

Risk: Competitors know who to poach; PR people spam us even more than usual.

-------

2) Show what we're working on. We already have internal wikis that are common scratch pads for teams working on projects. And most writers have their own thread-gathering processes, often online. Why no open them to all? Who knows, perhaps other people will have good ideas, too.

Upside: Tap the wisdom of crowds

Risk: Tip off competitors (although I'd argue that this would just as likely freeze them; after all the prior art would be obvious to all); Risks "scooping ourselves", robbing the final product of freshness. 

-------

3) "Process as Content"*. Why not share the reporting as it happens, uploading the text of each interview as soon as you can get it processed by your flat-world transcription service in India? (This may sound ridiculous, but it's exactly what wire services such as the AP have long done--they update their stories with each new fragment of information). After you've woven together enough of the threads to have a semi-coherent draft, why not ask your readers to help edit it? (We did it here, and it worked great). And while you're at it, let them write the headlines and subheads, not just for the site but also the punchier ones for the RSS feed and the one that has to work with the art for the magazine.

Upside: Open participation can make stories better--better researched, better thought through and deeper. It also can crowdsource some of the work of the copy desk and editors. And once the story is done and published, the participants have a sense of collective ownership that encourages them to spread the word.

Risk: Curating the process can quickly hit diminishing returns. Writers end up feeling like a cruise director, constantly trying to get people to participate. And all the other risks of the item above.  

-------

4) Privilege the crowd. Why not give comments equal status to the story they're commenting on? Why not publish all letters to the editor as they're submitted (we did that here), and let the readers vote on which are the best? We could promise to publish the top five each month, whether we like them or not: "Harness our tools of production! Make us print your words! Voting is Power!"

Upside: Maximizes participation.

Risk: If we don't deploy voting tools or (sigh) a login system, trolls may rule.

-------

5) Let readers decide what's best. We own Reddit, which (among other things) is a terrific way of measuring popularity. Why should we guess at which stories will be most popular and give those preferential treatment? Why not just measure what people really think and let statistics determine the hierarchy of the front page?

Upside: A front page that reflects reader interest better.

Risk: A more predictable and lowbrow front page.

-------

6) Wikifiy everything. The realities of publishing is that at some point you push the publish button. In the traditional world, that's the end of the story. It is a snapshot in time, as good as we could make it but inevitably imperfect. The errors (and all articles have them) are a mix of commission and omission--we hope for the best yet brace ourselves for the worst. But what if we published every story on a wiki platform, so they could evolve over time, just like Wikipedia itself? The original story would be the foundation of what could eventually become a version expanded and updated by readers (our Fortune 500 blogging wiki was an experiment in this). If you want to see the original version, just push the "original" button, or see any changes in-between by looking at the version history.

Upside: Stories live and grow, remaining relevant long after their original publication (at no cost to us!)

Risk: Stories get progressively less coherent as many cooks mess with them. Whatever brand authority the Wired name brings is diminished over time as the stories become less and less our own work.

----------------------------------------------------------------------------------

Needless to say, in all these cases I think the upsides outweigh the downsides. But if I'm wrong the consequences could be serious (Jason Calacanis's crowdsourcing experiment with the Netscape home page, which was something I admired, is nevertheless a cautionary tale. Although that was a portal in decline and clearly needed something radical to get traction again, the short term consequence of his Digg-like experiment was a significant decrease in traffic and a management turnover that included him). My strategy is to try these experiments in a limited way first so that if they don't work they won't take too much of the site down with them. But if they do work, we should be able to deploy them across the site quickly. And that would be very cool indeed.

UPDATE: I've clarified and amplified a few of these points here.

(*Credit goes to Evan Hansen for this term)

What would radical transparency mean for Wired? (Part 1)

One of the best things about 2006 was getting our website back so we could practice what we preach--now I bang on and on about "walking the talk". Turns out that between preaching and practice, preaching is by far the easier. The past six months have been consumed by porting 14 years of stories to a modern content management system, redesigning the site and submitting request tickets to an overworked engineering team. We have a pretty clear vision of what a magazine website in 2007 could be, but we've still got a long way to go. We're shooting for a February relaunch but it's already clear that some of the features we're most excited about will slip to later in the year. This happens. 

In the meantime, we brainstorm. It's been a fascinating exercise to think really broadly about what a media company today should be. We don't know (and neither does anyone else) but what's clear is that the landscape and reader expectations have changed dramatically since the mid-90s, when wired.com was first taking shape:

THEN: Bookmarks and habit drive traffic to the home page; site architecture and editorial hierarchy determines where readers go next. Portals rule.

NOW: Search and blog links drive readers to individual stories; they leave as quickly as they come. "De-portalization" rules.

-----------

THEN: Media as Lecture: we create content, you read it.

NOW: Media as Conversation: a total blur between traditional journalism, blogging and user comment/contributions.

-----------

THEN: Readers read HTML in a standard web browser window. If you want to be really fancy, design a whole new Flash interface that people will have to learn to get to your content. Charge for "premium content"? Sure!

NOW: More and more people read via RSS, where content is divorced from context. Media is atomized and microchunked. Even if readers do come to your site, the expectation is that the presentation will be a mix of HTML, AJAX, Flash multimedia and embedded third-party apps. Screens range from high-resolution wide displays to handhelds. Whatever you do, don't let your design interfere with web conventions--everything must be Google-crawlable and blogger permalinkable. Oh, and everything must be open and free.

-----------

THEN: We control the site. Editors are gatekeepers.

NOW: We share control with readers. Editors catalyze and curate conversations that happen as much "out there" as on our own site.  

---------

So there you have it, a motto for media in an age when consumers are in control:

"Catalyze and Curate"

In the next post, I'll talk about how that might work. 

Visualizing the rise of the Long Tail

In an article on "deportalization" (a theme I've been banging on for in my media speeches a while) Keith Teare offers a neat analogy for the shift from a few hits to many non-hits. He's referring to websites, but the same could be true for almost any products. 

Then:

Now:

Soon:

(Special note to Richard Florida: think of this when you're considering whether the world is becoming less "spiky", not more)