
Live with Vinton Cerf!
On the final day of the WebbyConnect 2007 conference, one of my many dreams came true. I got to meet and speak with Vinton Cerf, one of the key players in the creation of the modern internet, and currently Google’s Internet Evangelist. Tall, isn’t he? Judging from the photo I’m a little happy about this.
Vinton, or “Vint,” gave today’s keynote address, summarized below. He also attended the two morning sessions where he asked questions, spoke with admirers, and had a few choice words to say about SEO (search engine optimization).
For more photos from WebbyConnect, be sure to check out the WebbyConnect photos on Flickr.
And now, the morning sessions (all photos by Sandy Huffaker/Getty Images for The Webby Awards, available on Flickr):
The Art of Internet Judo
The first session’s panel included Rei Inamoto, Global Creative Director of AKQA, Amanda Kelso, Head of Production for Digital, Bartle Bogle Hegarty (BBH), and Nicke Bergstrom, Co-Founder and Creative Director of Farfar. David-Michel Davies, Executive Director of the Webby Awards was the moderator.
Rei Inamoto presented AKQA’s campaign for Halo 3. The challenge, Rei said, was how to engage fans over a 4-5 month period of time, from the time the campaign started to the launch of the game? One factor working in their favor was that gaming audiences are willing to go very deep and interact.
AKQA worked with heavily with the content writers from the Halo team, developing an ARG - alternate reality game - as marketing. This introduced an interactive narrative to tell the backstory of Halo.
There were five episodes created, each timed for release about one month apart. Each episode had a Challenge (phrased vaguely, in the form of a riddle to be solved), which led deeper content and clues. When the challenge was solved by 100 people, an “unknown truth” was unlocked to the general public. A unifying symbol was used throughout the various levels.
Some of the elements on this campaign included:
- Email to the Halo in-house list
- Coordinated live “protest” rallies that drew attention to the unifying symbol and pointed to the first step in the ‘challenge’
- A ‘landing’ web site with a hidden URL to be found, which led to
- A site with an audio file containing a secret password
- When the first 100 people solved the riddle and gave their passwords, a video containing more of the Halo backstory was unlocked to the public. The video contained clues to the next episode.
Fans quickly generated their own content around the Halo story, exponentially expanding the reach of the campaign:
- Wikipedia entries
- Other wiki sites
- Images from the video
- Translations of binary code included in the video
- Frame-by-frame analysis of the episode videos
Nicke Bergstrom presented Farfar’s “Heidies” campaign for Diesel. Nicke first showed a few background slides on Farfar. Farfar team members hard at work in costumes. The Farfar team accepting awards. Nicke about to jump off the roof of an outdoor Swedish sauna in wintertime, hoping to gain the newest phone from the Nokia executives he’s meeting with (he did).
Nicke hates the internet buzzwords, hates the advertising idea of “buying time.” Instead, he prefers the concept of “creating time” with the audience.
For Diesel, Farfar created the Heidies campaign to promote Diesel’s new underwear line. The campaign created two new internet personalities, “Heidi 1″ and “Heidi 2,” who enacted the scenario of kidnapping a Diesel employee in order to force Diesel executives to give them (the Heidies) a print ad, make them famous. The scenario was streamed live to the net via six cameras, and encouraged audience participation with:
- chat
- polls, i.e. “What should we do to become famous on the Internet?”
- Photos taken by the participants and posted live to Flickr
- Visitors affected the action by making suggestions and responding to polls.
The video was hosted on YouTube, then posted on Diesel.com. Since Diesel.com already had a high traffic volume, this leveraged that traffic to bump up the video’s visit numbers and rankings on YouTube. Overall, the campaign generated a huge fan response. Websites were created by fans to talk about it, customers came into stores asking for Heidies brand underwear (high retention).
And the kicker? Zero media dollars spent.
Amanda Kelso of BBH presented the Mentos “Trevor the Mentos Intern” campaign. She began by touching on how BBH spends a lot of time considering what a brand really means. Recently they’ve worked on Axe body spray and Levis. She mentioned their Smirnoff Raw Tea campaign, “Tea Partay,” which used white Hamptons residents rapping about tea to communicate the product with humor. There were quite a few ‘remixes’ of this video, generating broad audience response.
In the “Trevor” campaign, Trevor was given tasks by the general public, such as calling someone to sing them happy birthday in a heavy metal style. The public could communicate with him via
- chat
- calls
- email
- his online scheduling calendar
The interactive site included a live web cam on Trevor (he worked on-camera 8 hours a day), Flickr photos, and links to Trevor’s Facebook, MySpace, and blog.
Discovery: it turned out to be a huge amount of work to maintain the live site. Many people helped Trevor moderate the comments, emails, etc. Staff produced v-films of memorable sequences and posted them.
Unexpected effects: the member of the site D-Listed decided to bombard the site. Various techniques were tried to counter the attack, such as changing technologies, but nothing worked well enough. Finally, Trevor put up an “I Love DListed” sign visibly in his office space. This ended the attack.
Learning: Brands must learn to expect to be hacked or attacked, and embrace, rather than repel, any negative feedback. Be prepared with a response.
Audience Question: RE Halo, why go after those 100 people?
Rei: because they are the leaders / influencers. They actually build the brand.
Audience Question: RE Mentos - who was the audience?
Amanda: Teens & young adults.
Amanda, Rei, and Nicke: we had no media spend, but the other elements of these campaigns were very expensive.
Amanda: These technologies aren’t inexpensive, you spend a lot to set it up.
Rei: In 2005 the buzzword was “Viral.” In 2006 it was “User-generated.” In 2007 it’s been “Web 2.0.” Some clients and agencies think these are cheap methods, but just because it’s cheap doesn’t mean you should do it. What are you trying to say to the audience?
Nicke: Clients often think the buzzwords are magic. They try it and when it doesn’t do the job they say, “yeah, this Internet doesn’t work!”
Audience Question: was there any advance testing involved?
Amanda: Lots of scenario planning in advance, but no test group. For other campaigns BBH does use an ambassador network.
Rei: It’s important to always have Plan B. People hack the server, post unwanted content, etc. Sometimes Plan B must be created on the fly.
Question: Where do you draw the line between too edgy versus too conservative?
Nicke: Retouched ads are more annoying than the live site, more spammy and worse for young girls.
Amanda: There are of course liabilities for the brand, the ramifications must be thought about and the brand protected.
Question: How do you deal with a PR attack?
Rei: Be clear about your intent - to create, to entertain.
Question: are you monitoring online buzz about the campaigns?
Amanda: Yes, you have to, and you have to respond to it live. See MarketingPilgrim for free tools for buzz monitoring.
Nicke: To clear out annoying people we would shut down the entire site for 5 seconds or so. The unwanted people would go away. But trying to prevent them with ground rules, changing what they typed, etc., just provoked them further.
Question: what about costs for these campaigns?
Amanda: for the Mentos campaign, less than any site they’ve worked on in the past 2-3 years.
Nicke: the streaming was expensive, but the campaign was perhaps 1/10 the cost of other campaigns they’ve done.
Question (from Vinton Cerf): any spillover from other media?
Amanda: yes, we worked with a PR firm.
Vinton: so, there was a recursive effect. By using YouTube, other advertisers are essentially paying for your advertising. What was the concrete payback?
Rei: for Halo 3 had $170 million in sales in the first 24 hours of availability. It was the biggest game launch in history.
Amanda: the Mentos campaign just ended, so we’re still waiting for full results. However, on the older Diet Coke and Mentos project Mentos saw a 20% increase in sales in the U.S. last year, which appears to be a direct correlation.
Nicke: I wish the campaign had also been offline, had an offline component. It did not spread to people who weren’t already somewhat interested.
Throwback Blogging: PR Still Works
This session’s panel included Aileen Budrow of MTV Networks, Michael Maslansky of Luntz, Maslansky Strategic Research, and Gerry Byrne of Fidelis Global Management. The moderator was Jeff Stern of the Daily Reel.
(many apologies - I ran out of steam after that last session and took very few notes until Vinton Cerf answered a floor question -sn)
Audience question: Do you include Search in your PR strategies? Do you optimize for search?
Answer from Vinton Cerf: Let me just say this: Search Engine Optimization does not work! All those people trying to optimize their content are interfering with our efforts to deliver relevant results.
(Vinton then hinted at secret algorithms which, like the IRS checking random tax returns, could tag sites or pages for closer review.) “And, of course, the IRS doesn’t reveal its checking algorithms because that would enable people to fly under the radar.”
(I followed up with Vinton after the session, in the hallway. When I mentioned that one of my company’s services was search optimization and that we have, indeed, seen SEO work, he clarified: “What I’m talking about is people who generate thousands of sites just to link to their site and manipulate their search rankings, that kind of thing.” He radiated a kind of righteous exasperation about these kinds of attempts, repeating that they required so much time to deal with that it distracted effort away from improving result relevancy. You could tell he views this from an engineer’s perspective.)
Keynote Address: Vinton Cerf
With apologies for “not being an advertiser or a businessman,” Vinton began with an overview of the internet penetration world wide:
- Approximately 3.6 billion telephone terminal, including 2.5 billion mobiles
Samples of internet user distribution and penetration rate per area:
- Asia has 436.8 million internet users, but only 11.8% penetration rate. Asia has the most potential upside for internet expansion.
- Europe has 321.0 million internet users, at 39.8% penetration. Significant upside potential here, as well.
- North America has 232.7 million internet users, but has a high 69.5% penetration. Not much upside left here.
- Latin/Central America has 110.0 million internet users at 19.8% penetration
- Middle East has 19.5 million internet users at 10.1% penetration
Vinton showed a diagram of the original Arpanet, compared with a 1999 map of the internet, illustrating the explosive growth since 1969.
Socio-Economic Effects of the Internet
- Consumers becoming producers; the contribution barrier is practically zero
- Innovation
- Social networking
- Game playing, and the confluence of virtual and real worlds (example: People paying real money for Second Life services and the reverse).
- New business models, i.e. eBay, iTunes, Google, Amazon, VoIP, etc.
- Group interactions possible, not simply one-to-one
- Listen, play, print, view - all possible over the same protocol
Today: the physical implementation remains expensive. Very high cost for Google’s physical architecture. But less expensive than it would’ve been in, say, 1979; in 1979 three terabytes of memory would’ve cost $100 million. Today it’s only $2,000 - $3,000. The biggest challenge remains finding enough electrical power to run the physical aspect.
IP on Everything
(see the IP on Everything t-shirt debuted in 1992 here -sn)
The goals for the internet were:
- the network would remain data-agnostic
- packets could be sent by any system
- the web wouldn’t know what it was transmitting
- future-proofing, so IP (Internet Protocol) carries anything digital
Other parts of the world have a much higher possible bandwidth than the U.S. For instance, in Sweden you can get a 40GB dark fiber connection. Other countries such as Japan, Hong Kong, and France offer (or shortly will offer) similar capacities. For various reasons, the U.S. is far behind other countries in this.
We’re looking at a potential exhaustion of the current IP address version, running out of address spaces around 2010 or 2011. (see the IP Report site for stats and charts on this -sn) Next up to deal with this: IP version 6.
Legitimate reasons for ‘discriminatory’ (unequal) access to the internet include:
- game playing - games may need low-latency, so this data possibly should be tagged for this
- malware or virus filtering
- bandwidth - charge more for greater bandwidth
However, trying to prevent anti-competitive use of the underlying transport services, allowing access for users to all services on the web.
Vinton displayed a graph (above, photo from the WebbyConnect Flickr area) generated by Sandvine showing a deep packet inspection data display. This shows the various types of packets being transmitted across the internet during a period of time. YouTube takes up a very visible portion of the available bandwidth each day.
Key elements driving internet-media interaction include the possible transmission rate and the available storage space.
- If transmission rates are low and available storage is low, nothing works well.
- If transmission rates are high but available storage is low, then realtime and pre-recorded media can be streamed easily
- If transmission rates are high and storage is high, then realtime media can be streamed OR stored, and the same for pre-recorded media. In fact, pre-recorded media can be delivered to the end user at faster than real-time rates. Vinton mentioned that this is a largely unexplored space of opportunity for the video world.
IPTV
IPTV is the television model becoming the internet video model. With differences. IPTV allows:
- streaming and downloading
- mixing of all media together as IP packets. For instance, video delivered along with interactive elements or other tie-ins for additional functionality or promotional activity. The elements could contain ‘activate-able’ objects which could be branded, allowing users to decide what they want to look at further. This would pre-qualify customers via their expression of interest.
- multiple streams to multiple screens or speakers in a particular area (a house, for instance). “That’s the beauty of packet switching.”)
Mobility and Mobiles
There are over 2.5 billion mobile devices and counting. This explosion will create new:
- ways of payment. For instance, using mobile plan minutes to pay for purchases.
- challenges regarding the physical interface
- uses for mobile devices, such as becoming a centralized hub of connection and interaction in place of a laptop device. The mobile device would create a local network between interactive devices such as a keyboard, screen, and the internet, for example.
- More geo-location based services. “Transforming, huge opportunity as more and more geo-indexed information becomes monetized.”
- Navigation systems of various types
What’s next:
We’ll see more and more internet-enabled devices. (Vinton showed some existing examples, including fridges, photo frames, automobiles, even surfboards with embedded web access. -sn) The devices will be programmable using Java, Python, etc. Coming up we’ll see items like wine corks, clothing, and universal remote controls being internet-enabled. Imagine controlling your devices from anywhere! (except if you get hacked by the neighbor kid).
Challenges:
Vinton touched on three major challenges:
- intellectual property treatment. Perhaps the current model needs to be adapted to the new reality.
- Complex objects that can be rendered or interpreted only by a computer. For instance,
- 3-D interactive objects
- complex spreadsheets
- interactive environments
- Bit rot! This is by far the biggest challenge. As systems and software changes, digital objects will be lost. Imagine, for instance, trying to read a 1997 Powerpoint file in Office 3000.
- We need to somehow preserve the interpretive programs to read the objects.
- and the Operating Systems that run the interpretive programs
- and the hardware that runs the O/S’s
- For thousands of years!
- We have NO practices in place to preserve these objects and keep them accessible and usable. There’s a very serious technical problem trying to preserve information over long periods of time.
Question: can you expand on the potential for bandwidth growth in the U.S.?
Vinton: There is evidently very little incentive here in the U.S. to provide high-bandwidth access like other countries have. There’s not very much competition for broadband services. The availability and use stats that are reported are probably not very accurate. Wireless possibilities are still just possibilities.
Question: How do we design sites for the future?
Vinton:
- Use standard unicode character sets.
- Expand or find resources that can localize content to other cultures and languages.
- Plan for mobile access. A significant portion of the remaining Earth citizens will encounter the internet for the first time via a mobile device.
- Plan for the 2008 internationalized domain names - domain names in non-latin alphabets/ character sets.
Question: are the problems in the U.S. over bandwidth caused by old-timers in power?
Vinton: Yes, most politician rely on old business models. One solution may be simply generational change. However, now newer business models have proof behind them to convince.
Politicians first thought of the internet as only an outbound medium - they were surprised by the inbound responses of their constituency. Those politicians who miss the importance of listening and paying attention to and responding to this input will suffer. We’re starting to see some change in this area.
Question: Tell us about your work on the Interplanetary Internet.
Vinton: I’m working with JPL (Jet Propulsion Laboratories) on protocols for interplanetary communications. We’ve been testing in various terrestrial locations, solving such problems as delay and disruption of communications caused, for instance, by planetary rotations. We’re now confident we have solid protocols and plan to test them:
- 2009 - test new protocols on the International Space Station
- 2011 - test on the Deep Impact project
- Further propagation of the network, protocols and standards mission by mission to gradually build a solid, dependable communications system in space.
Question: Google has grown so large and dominant, are there any internal checks and balances in place?
Vinton: That question contains a hidden assumption that Google is bent on this type of ‘domination’, which is not true. (further comments to this effect.)
(Vinton concluded his talk at this point -sn)
View related topics: vinton cerf, webby connect
vinton cerf, webby connect