Sunday, March 22, 2009

Why Advertising is (NOT) Failing on the 'Net

Really interesting and controversial Techcrunch-guest-piece by Eric Clemons, Professor of Operations and Information Management at The Wharton School of the University of Pennsylvania. He makes the basic argument that users will ignore and/or disintermediate advertising online and therefore advertising will ultimatly fail. 
The internet is about freedom, and I suspect that a truly free population will not be held captive and forced to watch ads.
I think he's wrong. Here's why: 
  1. Performance vs Brand: He totally ignores brand advertising. Online, performance based ads are high-profile, and has dominated over brand advertising. This is due to the medium - the Internet is data rich and user-interactive (unlike TV that is passively consumed)
  2. Brand advertising isn't about conversions exclusively, it's about moving people through the purchase decision process and increasingly brand familiarity/ favorability with the goal of increasing the probability that someone will buy your product vs a competitor's. The point: an ad on a screen doesn't need to be clicked on to be impactful (from the point of view of a brand advertiser). Eg. If the Watchmen movie is launching this weekend, the studio would prefer that you click on the banner ad and explore the content THEN go watch the movie, but if you just decide to watch the movie (because the ad made the movie top of mind), they're also happy. 
  3. Everything he said about the Internet is true about meatspace TV advertising: people do not trust, want, or need advertising. He then argues that synchronicity of TV makes those ads more impactful, but has he never made a bag of popcorn during commercial breaks on TV?
  4. He narrowly defines "advertising" as the banner ad: corporate/ product sites are a form of advertising, and research shows that though this isn't the most trustworthy source, it's a primary source used by consumers to help make a purchase decision. 
Advertising on the internet suffers for a couple reasons: 
  1. Supply: Unlike TV where the advertising is "hardcoded" into the content, and there's a limited # of impressions possible, the Internet's has billions-billions-trillions of impressions everyday. At AdTech, interactive folks complained that the price of a broadcast impression is higher than an online impression. My arguement back: increase the quality of your content to have parity with TV and reduce the number of available of impressions/ user to the same as TV (say 5/ hour), and you have an arguement. 
  2. Relevance/ Trust: Relevance is really just a response to the supply problem. Mobile advertising,behavioral targeting capabilities, really rich niche content, social advertising really just allow advertisers to increase the probability of conversion (or a high-quality impression) and efficiency of spend. They're willing to pay more for inventory that "works" better.
In other words, I sense online people have high expectations for both the performance of ads and how they expect people will interact with ads. In general, people do not want to interact with ads, but they don't mind high-quality content sponsored by high-quality, relevant advertisers.

The problem from the publisher perspective - it's a race to the bottom, the market is too efficient:
  • Any high-quality content they make, is competing with "good enough" or "more relevant" content from another publisher. Eg I don't read CNN's political coverage, I go to huffingtonpost - I don't read their tech news, I go to Techcrunch. 
  • The Internet evens the playing field to attract ad dollars (not quite, but it's moving this way). Programmatic ad buying/ placement/ measurement means a big brand can just as easily advertise on a niche site as on a large publisher's site and if it performs, there's no reason to buy elsewhere. 
Both of these factors result in an ever-reducing ad rates: good for advertisers because it provides the opportunity for huge reach, bad for publishers because they have to actually provide good content. 

Saturday, March 21, 2009

Typealyzer.com - Analyzes Your Blog for Personality Disposition

Found typealyzer at thoughtgadgets.com. It's an interesting tool that scans the contents of your blog to determine your meyers-briggs personality bias. 



Monday, March 16, 2009

Bruce Sterling Sessions

Social networking bigger than email.
These artifacts are what we call books... Let me explain how these devices work. 
Publishing, in my lifetime, has never been in such a perilious state. 
Kindle is like a cassette plug-in for an atari. 
Less conerned about "death of author" more concerned with "death of audience"
Connectivity will be a signifier of poverty. 
Canadians are like back bacon, great white north. 
Ultra-right bookstore in Austin: Brave New Books
A year ago, video game conference, with ike refugees in the next room, who looked more futuristic? The greying game programmers?
The elderly will be the backbone of the social web. 

Old Man Nielsen vs. New Market Research

Daniel Neely   Founder/CEO,   Networked Insights 
Dave McClure   Troublemaker,   Founders Fund 
Jim Schroer   Founder,   EngageNextGen LLC 
Michael J Lambie   Digital Research Dir-,   Nielsen Company 

Neely (Networked Insights):
  • Social media has made access to data easy/ cheap. 
  • The same, companies are still selling things, need to stay relevant.
  • How to take customer message and put into my message
  • "Engagement" and "impression" are over-used words.
  • Old way: lots of data -> new way: actionable insights (free -> insight -> valuable insight -> actionable)
  • 2 by 2: high/low engagement, negative/positive sentiment. high/neg = pr, high/high = marketing
  • Tools: networked insights, visible tech, quantcast, radian, kissmetrics, trackur, scountlabs, comscore, nielsen, whostalking.com, google alerts/search, twitter. 
  • New way: actionable insights: adding anthropological perspective to the numbers
Measurement
Nielsen (Lambie)
  • Tools: Ethnolgraphy, neuro science, cams
Jim (Engage next gen)
  • Thesis: actionable/buzz means more than exposure.  Eg Teleflora - winner on superbowl (1400% inrease in "buzz"). Coke "buzz' went down post superbowl. 
  • Is social media a good proxy for word of mouth? If so, then it should impact spending. 
Agencies hide behind metrics (according to a brand). Not just sentiment, but interactions -> impacts marketing spend. 

Promoter score (McClure) don't promote until sentiment is positive. 
  • User voice, get satisfaction
Other tools (neely): truecast (invisible technologies), trackr, biz360
Mcclure: compete, hitwise

Browser Wars III: The Platform Wins

Arun Ranganathan  Mozilla 
Chris Wilson   Web Platform Architect,   Microsoft 
Brendan Eich   CTO,   Mozilla Foundation 
Charles McCathieNevile   Chief Standards Officer,   Opera Software 
Darin Fisher   Software Engineer,   Google

What's in it for Google in the browser game (to Google). You chose webkit. 
  • Initially, focus was "how to make Firefox more successful"
  • 2006, looking at other opps to radically improve browsers (multi-process, rending engines, make it faster). 
  • Gecko vs webkit - webkit was fast (surprisingly, fast). Code base was simpler and was adopted by mobile. Smaller footprint. 
  • Gecko is a full platform for app dev, webkit is only a rending engine. 
With the declining marketshare of IE6 (cheers), potential adoption of IE8 - there is no real majority rendering engine. Where to collab, where to compete? Chris (MSFT) Standards in the context of silverlight - you're the chair of the HTML working-group. 
  • Doesn't work, or evangelize siverlight, but there are scenarios where it makes sense. 
Standardization of web APIs - opera has a lot of skin in the game. 
  • There are some scenarios, but limits reach. If you want broad reach, you'll need to rely on standards. 
People still have to pull hair out to dev for web. What gives? (Mozilla)
  • Silverlight isn't taking over the web. A lot less worried this year, than last. 
How standards are actually made. HTML5 discussions: APIs, video, 2/3d graphics. Modifying the spec license. 
  • (MSFT) Distiguish between open source licenses: do whatever vs do and contribute back. 
Javascript performance wars: 
  • IE8 seriously by MSFT? Yes. Performance, though, not just focused on JS (holistic approach).
Battery dying...

Shift happens: moving from words to picture

Sunni Brown   Owner,   BrightSpot Info Design 
Tom Crawford   CEO,   VizThink 
Dave Gray   Chairman,   XPLANE 
Lee LeFever   Principal,   Common Craft 
Dan Roam   President,   Digital Roam Inc 

I caught the tail end of this session - I know Lee LeFever from my days in Seattle. His work rocks

Dan Roam (I think) showed some really interesting visualizations of the stimulus package. You can see some of them here

Beyond Aggregation - Finding the Web's Best Content

Marshall Kirkpatrick   VP Content Dev,   ReadWriteWeb 
Louis Gray   Author/Publisher,   louisgray.com 
Gabe Rivera   Founder/CEO,   Techmeme 
Melanie Baker   Community Mgr,   AideRSS Inc 
Micah Baldwin   VP Business Dev,   Lijit Networks Inc 

Postrank leverages "engagement metrics" to determine hot stories. 
Postrank , yahoo!pipes

Layers necessary for aggregation: 
  • Layer 1: enough data
  • Layer 2: enough meta-data and linking


Sunday, March 15, 2009

2 More Days at SXSW

Some post-Salt Lick recovery was required today and I'm refusing to carry around my lenovo T61p. Not only is my laptop huge, but I'm a dork among dorks - at least everyone else has sweet MacBooks.

The Seven Rules for Great Application Design
In this lively and interactive session, Robert Hoekman, Jr., the author of 'Designing the Obvious' and 'Designing the Moment', uses the audience to reveal the 7 essential design principles for achieving great application design and the psychology behind them. And he does it all without a single bullet point (gasp!).
Panelist(s): Robert Hoekman Jr (CEO/Principal Experience Designer, Miskeeto LLC)
  • People are bad at predicting their own behavior - ask users what they want, then ignore them. 
  • 7 Factors: 
  1. Understand users, then ignore them. Understand how users actually behave.
  2. Build only what's necessary.
  3. Support user's mental model. Eg. Squidoo calls pages "lenses" - users don't know what that is.  
  4. Turn beginner users into intermediate users ASAP. Eg. Get users registered and contributing ASAP. 
  5. Prevent errors. Eg. Don't throw random, unintelligible error messages at users. Just prevent it altogether and fail gracefully. 
  6. Design for uniformity, consistency, meaning. Make navigation/ layout, as consistent as possible
  7. Reduce, reduce, reduce (refine).
Emerging Trends in Mobile
Panelist(s): Rob Gonda (Dir of Mktg Strategy & Analysis, Sapient), Juan-Carlos Morales (Creative Dir, Sapient Interactive)

  • Nothing too insightful - some really amazing video footage of augmented reality
  • Some rehasing of Apple Appstore app download stats: 15k apps, 500m downloads, $1bn in revenue, 93% have apps, $25bn in mobile content downloads. 
  • Monkeyball is taking $15k/ mo
  • Allegedly, there's some announcement from MIT where they charge batteries in less than 7 seconds
  • Update (3/16): Good live stream here
Zappos - Tony Hsieh, CEO Zappos
I have to say, I'm not big on "corporate culture" but this guy really has it down to a science. Zappos has some pretty radical practices: pays new hires $2000 to walk away (to test commitment) and they publish an annual employee contributed "culture of Zappos". If you're ever in Vegas, they'll pick you up from the airport and will give you a company tour - these guys are serious about culture. 

His basic premise: brand identity comes when you get the culture right. 6 criteria for developing a healthy culture. 
  1. Decide if you want to build a long term brand
  2. Figure out your values and culture
  3. Vision, what you're thinking, and think bigger. Chase vision, not money. 
  4. Build relationships, not networks.
  5. Build your team: hire slowly, fire quickly
  6. Think long term. How people remain happy/ engaged: perceived happiness, progress, connectedness, vision. 
He got a little philosophical - models of happiness: 
  • Please -> engagement -> meaning
  • Job -> career -> calling

Friday, March 13, 2009

SXSW (Twitter) Tools

Twitterati indeed - it's almost comical. People are Twitter-crazy here here at SXSW.  There's so much data, I'm having difficult making something meaningful from the stream. That said, lot's of really interesting tools people are using. 

Twhirl (http://www.twhirl.org/): A Twitter Air client for your desktop. Looking around at people on their laptops, this seems to be the client of choice. The first question from the first session was "what's the hash for this session?". Twhirl allows you to "search" for hashes (eg. #swsw) and it automatially feeds the stream - really great during the sessions. 

Peoplebrowsr (http://www.peoplebrowsr.com). A very feature-rich web-based twitter client. What I really like about it is that you can "search" for and watch multiple streams concurrenting (eg I was watching #sxsw, #austin, #sxswi all at once today - it was pretty crazy). 

Outside.in (http://outside.in) This came up during Steve Johnson's (co-creator of outside.in) panel. It's actually primarily to be a crowd-sourced local-news site, but they have a clever mashup that overlays tweets, "story maps", containing geo-data on a map (ie. if you tweet "sxswguide: Heading of to Six Lounge for the Social Media Group party. sxswi (@hametner) MARCH 13" it will put "Six Lounge" on the map). 

Other tools people seem to be using: 

Sched.org - The SXSW schedule site sucks really bad. Even the printed schedule is unusable this year (I think last year it showed the sessions in order, with a brief summary - this year, not so much). Sched.org is a really usable web-based, SXSW-dedicated, calendar that lets you filter by show (interactive, film, etc) as well provide high-level summary of the sessions. It let's you find free booze and food, and let's you merge calendars with your fellow-attendees. 

happyhour.org - SXSW-dedicated party schedules. The sanctioned-sxsw schedule is better this year (it actually shows some of the parties), but this online tool is definitely more complete.

Social Engineering: Scam Your Way Into Anything or From Anybody

Brian Brushwood   Host,   Revision3/Scam School

Interesting, albeit comedic, session by Brian. He's the host of an InternetTV show hosted by Revision3. His session was basically a summary of his show: using influence concepts for magicshow-like entertainment (eg pick a card, choose a word and he guesses it). It was definitely entertaining. 

Check out the most recent show - he does a bar trick where he charges a straw and pretends he possesses telekenisis. 

That said, and he even referrenced it later in his session, a lot of the concepts are captured in the book Influence by Robert Cialdini (which I highly recommend). It was also somewhat similar to the concepts contained in Josh Porter's "Designing for the Social Web"

The Ecosystem of News

Steve Johnson from outside.in is the author of "everything bad is good for you" and "the invention of air". 

News = "old growth media" 

Back story:
  •  going to the magazine store 3-4x / day to get new issue of mac world in the 80's. 
  • Moved to nyc, joined compuserve for mac week. 
  • '93 - wired magazine launched.
  • macintouch '
  • salon - scott rosenberg
Now: engadget, 'technica, gruber, norman, pogue, mossberg, macrumors. 
Mcclewan vs wilson (now).
Blogging is inherently "parasitic" (bloggers link to traditoinal news articles) 
New growth: depth and surface
Outsidein - geo-twitter
Jeff jarvis - "do what you do best, and link to the rest".
Newspapers "paper" readership has gone down, but online has increased. There's still a market for "authoritative"/ "editorialized" content that newspapers bring. 
Ecosystem: professional, prof bloggers, non-profit journalists, amateur bloggers, diret events (eg obama race speech), public data/ apis -> commentary: pundits, bloggers, scholars ->curation: social media, professional/editors, aggregations, group filters -> Distribution: traditional, aggregators, viral world of mouth

Q's: someone needs to get paid to do this effectively. Business model? 
A's: no answers

Lots of interesting techs: 
sched.org
peoplebrowsr.com
outside.in



User Generated Content: State of the Union

In general - this panel wasn't so great. They didn't mention monetization models.

Moderator: Chris Tolles Ceo, TopixDean Mccall   Founder,   IdeaGin 
Stephen Newman   Chief Exec Officer,   Mouth Watering Media 
Todd Morrey  Mosso: The Rackspace Cloud 
Wes Wilson   Pres,   IncSpring (crowd souced brands/ logo)
Chris Tolles   Ceo,   Topix 

Incredible - the first question asked was "what's the twitter hash for this panel?" It was #UGC. 

Why UCG matters: In 1998 only geocities was UGC, now, 3 of top 10 (fox, wiki, facebook)

Good examples now: 
  • Wes: YouTube (based on ads), 
  • Todd:Twitter (tho no money!). Seems like an auction-like model seems to work - eg. Ebay. 
  • Newman: TV - think reality tv, america's funniest home videos. Etsy
  • Wilson: Linkedin ($4M in ad revenue), Threadless (people can upload content, and people can buy the crowd-created designs).
?Gracenote - part of itunes for track listing. 
We cheered to beat the room next door. 

Money off the back of the masses? How fair is that?
  • As a user, when you're not paying for a service - it's expected.
  • But future models will allow users and company to profit. 
  • There's more to it than $ - derivative products?
  • Facebook Ts&Cs - eg crowdsourcing terms.
How to prevent the riff-raff from destroying site? 
  • No programatic way.
Other descriptions for UGC - "community curated works"? No one knows what CCW is!

Slashdot (karma). 
Hierarchical - moderators can nominate other moderators. Wikipedia can see everything that's going. The more transparent, the less easy to game system. 

Craigslist - $100m/ year. $8M on digg in 3 quarters last year. 
Monetization is easier when closer to "the sale". Ie further down the purchase decision cycle. 

Granovetter thresholds mentioned on twitter - on wikipedia

Blog catelogue - bloggers unite - ralley around a topic. 

Not sure what any of this had to do with monetization of UGC.

SXSWi 2009 - Party Time, Don't Tell the Guys at RIM

First things first - I haven't written on this blog since Oct 2008. It's interesting how getting sucked into the corporate machine prevents that. I'll be at SXSW today to next Friday - hopefully I'll catch some interesting sessions and some cool music (J Tillman on Thursday night!). 

The lines are looking crazy for registration. I'd read that the "economic conditions" will result in reduced attendance. I'm sure the poeple waiting in the mile-long lines aren't feeling that way right now. I got here early, registered, and ate BBQ. 

Today's agenda: 
Not sure there's any party-time tonight, but everyone is talking about happyhour.org - a party listing iphone app, from what i can see. I'll check it out now.