Ross Mayfield's Weblog

Promoted Tweets and the Next Ad Recovery

A lifetime ago I worked on an Eastern European telecom group's IPO and one of my favorite slides in my deck was borrowed from Goldman Sachs to paint the picture of a Multimedia Cycle. Yes, I said multimedia, but right now there is something shifting in the ad market that may or may not mean a recovery to yet another boom for the internet.

Picture this in a cycle:

Ads -> Content -> Applications -> Telecom Services -> Telecom Infrastructure -> Repeat

In 2002 I described the multimedia cycle like this,

"Just as potential customer spending drives advertising expenditures, Ad spending drives the development of content to create ad space.  Development of content drives spending telecommunications services to delivery it.  Development of telecommunications services requires investment in applications and infrastructure."

I also pointed out how GDP is correlated with ad spend and noted that ad spending drives internet recovery.

In 2004 there were signs the ad market recovered and the Web 2.0 boom was getting started. Google drove advertising innovation. To the point where way too many bloggers thought they could live off it. This ad recovery fueled a venture boom and falsely led many startups to leverage someone else's business model instead of finding their own. Many of course didn't make it through the crash, and those that did focused on creating their own leverage.

Ad revenue isn't what it used to be. And it won't be. Which may lead you to have concerns about the current recovery:

• The last boom didn't deliver innovation in new ad formats and metrics, and there still isn't a solid ad model for social networks like Facebook. Put simply, advertising still works for the old web. More on this later...
• Ad commoditization is moving far beyond text ads into other formats. The display ad market has moved to Real Time Ad Exchanges (disclosure: I'm an advisor to Triggit, which helps helps buyers optimize real time spend). Conversion rates always decline as the audience becomes sensitized, with the recent exception to newer formats like video ads (won't last long.
• The B2B lead gen business has been cutthroat as well. 
• CPMs, CPCs, CPEs & CPLs are falling through the floor. In part this may reflect the broader economy, but every last efficiency is being wrung out of the markets.
• What this means is what we've known. That traffic & engagement are less easy to monetize than during the last recovery. Or at the very least less than what was believed at the time.

This may lead you to believe the multimedia cycle will not be fueled, and internet recovery will stall the creation of the next wave of startups. But things have changed.

In Apple's Garden of Eden, you can skip advertising and directly monetize your content and apps. Subscription and ad models are ready for the taking as well. But this forbidden fruit drives your app down to $.99 and you only get 2/3rds. You find yourself in the hit business, and usually it ends up being Hollywood's studio model instead of an indie.

Startups already have to innovate their own ad models, and rise above them. When shit lobsters (a complement) like ^Dave McClure are calling for coupons, that's what he means.

The big case in point is Twitter's launch of Promoted Tweets, the launch of Bill Gross' TweetUp and other new ad formats for the real time web like MyLike. It's not just about Twitter finally finding a business model, built upon the problem they created for brands. They are testing a new ad format and ranking algorithm that hopes to strike a balance between advertiser and user. And the real test is if they gain distribution through their search and other ecosystem partners starting tomorrow @chirp.

I'm going to go out like a bird on a limb here to suggest that it may be another tipping point for Twitter. One that involves actual tips. And supports a good portion of the ecosystem. This will serve as a litmus test for ad formats on social networks.

But the key for Twitter is if they can gain a distribution advantage for how they monetize. The qualms of them buying partners shouldn't concern partners, a lack of acquisitions should. "Filling holes" is a ridiculous analogy, btw. It reminds me about that story of a Dutch kid taking his finger to a dike. But not literally.

Others ad formats are arising, like with SlideShare's AdShare for professional sharing (advisor disclosure).The question I have is if new ad formats and other ad innovation will fuel another modest boom, because ads will always be a portion of new rising internet business models that are widely adopted. What form will they take? And will social gain a creative form that is valued as much as the creativity we have despite it.

The End of Embargoes

You gotta love Mike's style.  TechCrunch is banning embargoes.  Now, bloggers are already notorious for breaking them, and at least TechCrunch is formalizing it.  And by doing so, I think embargoes will gradually go the way of the DoDo.

The value of an embargo is the big launch, with multiple journalists covering the same thing (same message) at the same time.  Often, an exclusive is given to one of the journalists, say a customer interview.

As I've said before, we are moving from a world of exclusives to inclusives.  Inclusives build the story in public, involving more people in the conversation.  This carries an observer effect, leading to a more engaging story that explores more angles in more places.

Don't read the front page

David Weinberger highlights a comment by Michael Wolff in Vanity Faire about his new news site, Newser.com.

The metaphor, for 150 years ― from print to radio to network to cable ― has been the front page: important stuff first. "It should have to do now with falling through something, or floating through the totality of information or of intersecting worlds and interests," offers [Patrick] Spain, not a man wild with his metaphors. [VF, October, p. 126]

I've been saying for a while, and I think in Everything Is Miscellaneous, that the new front page is distributed across our day and our network. Much of it comes through our inbox. It consists of people we know and people we don't know recommending items for our interest...

When I was around 10 years old, I was a member of the Optimist Club.  The group leader was a great guy, ran a construction company and drove a swanky green Caddie. Happy and successful guy.  I'll always remember how he said to never read the front page of the newspaper first.  All it would do is bring you down about what's going on in the world.  He suggested instead to start with the funnies and work your way back to the front page.

I find the same thing to be true about my inbox, the modern front page as David suggests.

Advertising is not Democratic

I was disturbed to read a pandering post by a Google employee that decries Michael Moore's documentary Siko and offers advertising as a means for the U.S. health care industry.  Others were, and Google's official position that was no position.  Dan Farber has been following the story, and added this update:

Update 2: Now we have an explanation from Ms. Turner regarding how to read her post. She just meant to state Google’s position that “advertising is a very democratic and effective way to participate in a public dialogue.” I won’t argue with the idea of advertising as democratic. Anyone with the money or winning bid can get their message out into the ether.  But ads tend to be one-sided sales pitches without footnotes, not a public dialog. If we want a public dialog, having the two opposing sides in a public debate would be a far better way to educate the public.

I will argue with the idea of advertising as democratic.  It is the opposite.  Spending isn't speech.  Sure, U.S. health care can buy ads to be placed in context alongside public discourse.  But not everyone can.  It concerns me that the bright people at Google could be talking themselves into believing that either advertising is democracy, let alone that it helps democracy.

If the U.S. health care industry really wants to respond to Sicko, they will engage in, if not host, online communities for civic dialog.  However, most online communities these days are powered by advertising. Community hosts and ad networks have to balance against the very strong incentives to smudge context and placement until where the line between paid and unpaid content are blurred.  A balance is struck, not unlike between editorial and publishing in traditional media, but with a very big difference in that the audience has the choice to go elsewhere with a single click.  Or create their own without the influence of advertising. 

Transparent Media and Wikify Everything

Chris Anderson is on a roll about transparency  and six tactics for transparent media:

1. Show who we are.
2. Show what we're working on.
3. "Process as Content"
4. Privilege the crowd.
5. Let readers decide what's best.
6. Wikifiy everything.

An interesting thought experiment is not only would this work for Wired, or a media company, but would it work for any company.

Expounding upon the reactions to his post, Chris clarifies:

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.

Actually, an interesting question is if you could make the short form more like the long form.  Many a time have I been interviewed by a writer and then had a photo shoot with someone that had little or no idea what the article was about.  Photo editors don't give a lot of art direction, especially to freelancers.  Wired features are more collaborative than the norm, but I have to ask if tighter collaboration could be fostered for short form stories.

At Ziff Davis a while ago, this process between the art director, writers and editors was made transparent through a wiki:

In order to put a story together, Art, Production and Editorial departments have to work together. The Editorial team is on the 8th floor, the Art team is on 9th floor and someone is always out of the office.

The team and department heads realized they could use Socialtext for day-to-day coordination, scheduling and requests. The groups need to communicate to each other what's needed for stories in progress: Art, HTML requests, copy, all in a back-and-forth conversational style. The groups set up a page for each activity, posting requests as they come up, and tracking fulfillment of the request. A structured approval process wouldn't address the need for iterative communication. "Because its not always a regular process, we can communicate when something isn't clear and coordinate getting it done," said Kennedy.

But the interesting question to ask with any process, especially an ad hoc process, is what would happen if we increase the scale and transparency through a wiki?

Chris suggests this is possible by wikifying everything.  But I'd suggest that for most newsrooms there is a preliminary step of internal transparency, and then involving freelancers.

Pincus Punked by the Post

Mark Pincus was made an example by the Washington Post today because he stuck to his principles and refused to be censored.  The article is a classic mainstream media take on the abject horrors of newstream media and how it can impact reputations.  Mark's response details many of the inaccuracies.

The irony is now Mark's reputation is at stake because of a mainstream media piece that does no better than what the newstream would.

Guestblogging by Telegraph

I'm guestblogging for Shane Richmond at Telegraph.co.uk.  First post: share control to create value. 

Most of my posts will probably be relating what's readers of this blog already know.  But if I make good fun of British newspapers or come up with something unexpected, I'll cross-post here.

As an aside, it is already interesting to have photos chosen for your blog post and a headline writer to have his way, calling the first post Why it's good to loose control.  Perhaps I have.

Wiki Wired

UPDATE: Veni. Vidi. Wiki. The published story, and commentary by Ryan Singel, The Wiki That Edited Me.  Charts of editors and saves by Peter T.

I believe the Wired Wiki experiment can be called a success, and yesterday I would have said it was doomed. Just came back from Wiki Wednesday, where Wired reporter Ryan Singel held a conversation about it.  How we conducted the experiment, what part of the editorial process it was directed at it and the participation of the community gives us a lot to learn from.

Do recall that the use of wikis in journalism has been significantly tainted by the LA Times Wikitorial debacle.  It was a failure in wiki implementation, goal setting, content structure and moderation.  While the media has embraced public blogs, they still have a while to go before public wikis are accepted. 

When I was an intern I got into an argument with the editor of the Washington Post that ended with him telling me I have a perception problem.  Since then I have been trying to prove him wrong, but that's a larger story, right now I might tell him the same thing.

There are different parts of the editorial process where wikis are perfect fits.  Public wikis of course run the greatest risks, but I believe these risks can be managed with the right practices.  Up front, this experiment was different from Wikitorial:

• Monitoring tools like Recent Changes and History were made available to let the community moderate
• Requiring registration to edit made contributors more accountable
• It could be argued that a WYSIWYG editor enabled domain experts to contribute, but in this case the domain was wikis, so it probably wasn't a factor.  Nevertheless usability always matters.
• Besides the article, we use a weblog for submission of headline ideas and an included page for the deck
• Choosing an article, instead of an editorial, provided an implicit guideline for what was acceptable.
• Most importantly -- a clear goal for the collaboration was set

There was one lesson from the Wikitorial, something I saw coming back then, that we had to employ in the middle of the project.  With the Wikitorial, an edit war ensued with differing viewpoints.  Jimmy Wales stepped in and forked the page, creating pro and con editorials.  With a wiki, there is space for everyone, even when the topic is war.  In this, the topic was wikis where lots of people have a stake.  Wired endorsed an Enumeration to also be published, for remaindered links and references.  This served as a pressure release valve, to let the quality of the main article improve.

We made some mistakes, perhaps on purpose:

• The collaboration was for an open group of participants to play the role of editor to Ryan's original submission.  This is a very unusual role for most participants, and most chose to edit directly in ways that an editor accountable to both the institution and journalist would not.
• While there was a goal, there were no guidelines.  Closest thing was pointing to the Wired Style Guide.  There were no rules, norms or conventions.  Unlike the Esquire experiment, there was no benefit from Wikipedia's established guidelines.
• Leadership was passive.  Wired explictly did not edit the article and did not pass judgement.

The article initially evolved reasonably.  A pattern I have seen before, say with sharing Wikified Books, where most of the contributions were lightweight and adding in references and associations that came to the minds of readers.  Personally, I stayed out of the initial editing fray (Wired did too, more on that later) to leave room for others.  But a wave of edits came in.  Angela called it "here's my favorite wiki" links.  This did reflect the topic of the article, a domain of vendors, open source and public wiki project managers.  Almost every wiki vendor added a link to themselves and editing in their positioning.  Part of this is a sad commentary about the space, I can tell you that if this experiment was done 2 years ago, the contributions would be more, uh, wiki-like.

There were some real gems, particularly in the education section.  One person took it upon themselves to interview an expert at Harvard after coordinating with Ryan and contributed a quote that persisted.  Someone suggested an expert to Ryan on the Comments page, but he didn't have time to interview her.  She got word of the experiment and contributed persistent edits herself.  There was a minor dispute within the section, and some backchanneling between the parties, but the result stands.

At one point, someone stepped in made the first significant edit.  They were kind enough to leave the remainder on what was the beginning of the Enumeration page, with an explanation:

Whereas the Wired Wiki story became too long, and became a soapbox for too many wiki, editors moved some of it to this page.

Hark, and know that you are upon an epic enumeration of wiki. Here, mentions of many wiki sites shall you find, and links shall you encounter.

At first I thought it was crazy, but then saw the wisdom.  Unfortunately, this created a vacuum that was quickly filled with the same.  Yesterday, with the deadline approaching, I thought the experiment would be a failure.

This morning there was a significant amount of edits, including some multiple detailed edits by one editor.  Suddenly, the article was a story again.  And the edits persisted.

At present, there is no plans for a community and I wouldn't say that one took root.  But some common understanding was reached in a short period of time.  I gained greater confirmation about some mediative techniques and moderation practices.  There are better parts of the editorial process to apply this too, some proven already, and many happening behind the firewall, but much work to be done.

The result is a good story, dear reader (at the moment), I leave to you to judge.

Participatory Media at Wiki Wednesday

Tomorrow night's Wiki Wednesday Palo Alto will feature an open conversation with Wired News reporter Ryan Singel about lessons learned from the Wired Wiki experiment.  I expect we will cover the broader topic of participatory media.

We'll have a parallel conversation for developers about the REST API and STRUT (a conversion toolbox.

See you Wednesday night, 6-8pm, at Socialtext.

Between Popular and Personal there is Social

Every time I see Gabe Rivera of TechMeme, I ask for the same thing -- MeMeme.  Give me TechMeme where the core index is based on who I read, about 150 people at any given time, to show me what my friends are interested in.  I used to ask this from people who make Newsreaders.  Because simply somedays you are too busy to read everything, but you want to make sure you haven't missed something big.  That's the real value I derive from TechMeme today.  But what I really want find something that is big with my friends, which in the larger blogosphere is actually something small.

Today we have two new and seriously great kinds of attention tools.  Newsreaders give us the ability to personally personalize.  Combined with persistent query feeds, you can follow the people and things you know you want to read.  Similarly, social networking services with a purpose let you aggregate the objects of your friends, be it pictures with Flickr, or posts with Vox.  Tagging then lets you pivot for social discovery, but that is digging deeper than you often have time or interest for.

TechMeme and others show us mass popularization.  Different communities help things bubble up.  In Social Software, you first saw this with Blogdex and DayPop.  What in the blogosophere has the most attention within a given time period.  Now we have Digg, del.icio.us/popular, Reedit, Netscape, Technorati, YouTube, Dabble, Last.fm, Flickr Interestingness and a gazillion other increasingly rich examples.  This is a Wisdom of Crowds we couldn't gain before for discoverable knowledge.

As an aside, I wonder how original Slashdotters feel about Diggers' favor for a popular  answer rather than a leading question.

So more concisely, what I hope develops:
* Tools that let me personally personalize should give me just one more degree of interestingness and popularization. 
* Tools of mass popularization should give me social popularization

Since Flickr has both kinds of attention tools, let me give specific suggestions for extension.  For within My Contacts's Photos, show me the most viewed, favorited and commented by my contacts.  Then show me the most viewed, favorited and commented pictures by my contacts in Everyone's Photos.

Now, this is just one user's greedy suggestion, and there serious usability and algorithmic challenges to overcome.  But what I'm getting at is part of the future of media.

The other night I watched the evening local news broadcast for the first time in a while.  Its funny how local news attempts to localize national news.  The idea is that if they show you a Mom in the Bay Area of a Soldier in Iraq, you can relate to that and it brings the story home.  But unless the story originates from that Mom or Soldier, it is just an overlay with too much of a contextual shift.  Similarly, when an item of local news is made national -- it is too shallow for our local tastes and we are attracted to it simply because or fair city is made popular. 

I empathize with the expert editors behind these mass media and their attempts to connect the interesting for me, when me is lost in a demographic.  But I've had a taste of going direct.  When I carry the burden of discovery, and float around YouTube's popular and related clips, I can compose a broadcast for myself.  The outstanding political commentary, funny stuff and best soccer highlights from around the world.

But after a long day of work, I'm tired, and want the network to work for me.  Cue up not what is popular, or what the people I subscribed to produced.  Cue up what my social network has found interesting.  At any given time it may be local, national, international, topical or mundane.  Of course, in the process of actively consuming it, I'll leave behind breadcrumbs of attention to make it better for my friends.

UPDATE: I've been ranting about this for years.  Sam Ruby hacked together a nifty MeMeme and the result shows a clear and simple foci of attention (a post by Spoksky at last glance), FeedDemon has something in the works and Tailrank has something close.