Sam Pullara vs the “Google Killer”

Yahoo!’s Sam Pullara cracked me up yesterday. In about a day he replicated the user interface and functionality of the overhyped $33M “Google killer” startup Cuil.

Cuil released their service on Monday to much fanfare though not without some major glitches. Sam single-handedly brought the well-funded and overconfident company back down to earth with a few strokes of the keys.

He calls his creation Yuil. It uses Yahoo!’s recently launched BOSS search platform and looks and feels identical to Cuil. And the output of Sam’s creation appears to be at least on par with what they built.

“I thought I would have a little fun with it and put together a quick parody of it by mashing up their UI and Yahoo!’s search results. As usual, the biggest problems I had were related to my pathetic Python skills.”

A few lessons here:

  • A good feature does not necessarily make a good startup even if it costs $33M.
  • The market supports ambitious startups. Very good.
  • Ambitious startups can crap all over the market. Very bad.
  • Being a former Google employee does not in itself make you important.
  • Don’t mess with Sam Pullara.

UPDATE: Ironically, YUIL is now dead. I suspect he’s either rebuilding it somewhere to deal with the traffic load, or someone asked him to take it down.

Defining online media platforms

My thinking about what platforms and ecosystems look like in the online media world seems to evolve constantly. It has certainly become more clear, bigger and more nuanced the last 2 years or so, but the language to describe a media platform feels very unfinished still.

It may be that the word ‘platform’ is throwing off conflicting ideas of what exists in my head. If you ask 10 people to define a platform you’ll get 10 different answers.

You can think of a platform in several different ways. There’s the functional role it serves. You can talk about the ecosystem around it. Some explain it in terms of the pieces and how they interact. It can also be an abstract concept or more of a strategic view of things.

Blogger (and conspiracy theorist) Kid Mercury has a paper of sorts defining platforms in terms of the strategies that they serve and their relationship to products, particularly in the Web 2.0 context:

“Products are “things” (goods, services, experiences, etc) that are sold; platforms are the intermediaries that deliver products. True to Yin/Yang form, the two are complementary yet opposing as well; platforms cannot exist without products, and products need platforms to be put into context and to be found.”

Wikipedia has several entries on platforms including “computing platforms“, “political platforms” and “platform shoes“:

Platform often describes the set of hardware components that make up the computer itself, that the software is written to target (often just described as “written for an architecture”).”

Auto manufacturers initiated platform strategies in the 1960’s and ’70’s to improve several aspects of their processes:

“Vehicle platform-sharing combined with advanced and flexible-manufacturing technology enables automakers to sharply reduce product development and changeover times, while modular design and assembly allow building a greater variety of vehicles from one basic set of engineered components. Many vendors refer to this as product or vehicle architecture.”

Similarly, John Hagel and John Seely Brown noted how platform design in auto manufacturing in Asia has enabled amazing efficiencies in their 2006 paper “Connecting Globalization & Innovation: Some Contrarian Perspectives“:

“Honda’s share of Vietnam’s motorcycle market, for instance, dropped from nearly 90 percent in 1997 to 30 percent in 2002. Japanese companies complain about the “stealing” of their designs, but the Chinese have redefined product architectures in ways that go well beyond copying, by encouraging significant local innovation at the component and subsystem level. “

David S. Evans, Andrei Hagiu and Richard Schmalensee go into great detail about various approaches to platforms taken by companies like Microsoft and Apple in their book “How Software Platforms Drive Innovation and Transform Industries.” They explained several models for how platform businesses work:

“In a multisided strategy, the software platform mainly facilitates inter-actions between the sides of the platform (particularly applications vendors and end users). In a single-sided (or merchant) strategy, the platform either produces the complementary products itself or buys them and resells them to end users. “

The book jumps from strategy down to specific detail such as this bit about how to develop relationships with customers:

“In the case of the software platforms we have examined, even the weakest relationships are far deeper than the arm’s-length relationships one sees in many one-sided industries. Software platforms can’t have direct relationships with the thousands of small developers, hardware makers, and peripheral device makers. Yet they document and make APIs available to developers, provide interface information to hardware and peripheral makers, and make sure their platforms have the relevant drivers for the peripherals. And they develop relationships through large developer conferences and small focus groups that bring some of these smaller players together. At the other extreme, software platforms often have deep relationships with several larger partners. These relationships involve regular exchange of information and joint work on defining new standards and specifications. They may also involve joint investments in product development or marketing.”

Thomas Eisenmann of Harvard Business Review has done some interesting work addressing the network effects of the modern media platform in his paper called “Platform-Mediated Networks: Definitions and Core Concepts“. He uses Visa’s and Microsoft’s XBox platforms to compare and contrast different methods for creating network effects. He talks about how a jointly sponsored platform like Visa’s leverages more complicated but more scalable relationships vs the single-proprietor platform like Microsoft’s which has lots of dependencies.

Visa vs Xbox platform ecosystem
He builds on this view in his follow up work with Geoffrey Parker and Marshall Van Alstyne for MIT. In “Network Platforms – Core Concepts” he builds network effects as a necessary component of a platform:

“A “Network platform” is defined by the subset of components used in common across
a suite of products (Boudreau, 2006) that also exhibit network effects. Value is exchanged among a triangular set of relationships including users, component suppliers (co-developers), and platform firms.”

They spell out some of the trickier issues the platform organizations need to consider such as channel conflict:

“Platform providers must determine how much of the value created through network
interactions they should seek to capture and from which users. consider who adds the most value. A bigger network served by a single platform can create more value in aggregate, but users may worry that a dominant platform provider will extract too much value. Likewise, when the participation of a few large users is crucial for mobilizing a network (e.g., movie studios vis-a-vis new DVD formats), conflict over the division of value between platform providers and “marquee” users is common.need to give in order to get. Controlling most of a multi-billion dollar business is better than controlling all of a million dollar business. “

Just having a platform and a platform strategy is not enough. There are organizational requirements that make it possible to drive a platform business. Kid Mercury adds some insight into the ways in which organizations must institutionalize capability creation to serve platforms rather than functional hierarchy to serve products:

“There can be no long value chains where each employee is a rung in a ladder, with all the value ultimately flowing to the top. Such hierarchical organizations are essentially immobile by design; they are not capable of creating new capabilities because everyone in the vertical hierarchy is participating in a way that only serves the existing value chain. This is great for incremental innovations, as such a structure essentially institutionalizes the process of adding more value to existing value chains. It is not so effective, though, for creating new value chains.”

I’m looking for more views on platform strategies and ecosystems, so please comment below if you have any favorites. I haven’t found as much dialog and literature as I’d like.

I suppose all this goes hand-in-hand with some of the discussion around designing for growth. Maybe I’m alone here, but I’m more interested than ever before in this old magazine idea.

Chad Dickerson to join Etsy in New York

Chad Dickerson writes about leaving Yahoo! and starting his new job at Etsy running their technology efforts. He’ll be relocating to New York after spending 10 years in the Bay Area.

“To my Bay Area friends and colleagues, you have given me so much in my time here, I can’t thank you enough. The time I spent out in California this past ten years has literally been life-changing. To my New York friends, I look forward to reconnecting.”

It’s great to see talented people getting rewarded for their hard work. Unfortunately, this is a loss for Yahoo! who needs forward-thinking leaders like Chad who can make things happen. Retention must be top of mind at Yahoo! before key institutional knowledge slips out the door and forces people to rethink things that have already been thought through.

There are lots of great reasons to participate in the future of Yahoo! where the Open Strategy stuff is unfolding. The Flickr Era set the stage for a lot of these smart ideas at Yahoo!. I only worry that the pace of release at the company will fail to create the impact that will make those changes matter. It’s not uncommon for great technology to lose due to bad timing.

The timing worked well for Chad and Etsy, though. He’s going to be a huge asset to Etsy at an important stage in its growth. If it wasn’t already, Etsy should be ranked high on everyone’s companies-to-watch leaderboard.

The Flickr Era

I’ve been wondering how the Flickr era was going to close for a while, but the resignation letter couldn’t have been a more fitting ending to the story. Looking back, I think there were a lot of smart things that Yahoo! did during that time which are easy to overlook. The end result of the action muted the intent, in many cases, but the desire to do the right thing was and is still very prevalent.

I’ve been playing around with Swivel, Dipity, Timeline and Google Docs a bit recently to try different ways to use data for storytelling. So, the flickr departure news seemed like a great time to give these things a shot.

Below is the “flickr era timeline”, the smart things that Yahoo! did from early 2005 to the middle of 2008:

I’ve also opened sharing on the spreadsheet. Feel free to add to it.

Here’s the embed code for the timeline:

London hackers are back

It seems very strange blogging about this topic from my desk here in London. If there’s one constant in the Internet, it is change. But one thing that is still the same a year on — hacking is alive and well.

This coming Saturday hackers will again descend onto London’s Alexandra Palace to come up with some new ideas with their fellow coders. The same basic rules apply: hack anything you can think of in 24 hours and then demo your creation in 90 seconds. From what I hear, there are a few new twists to the event that could make it really interesting.

BBC is driving the event this time, and they’ve called it Mashed. The Guardian is one of the sponsors, and they’ve just opened up an additional 100 tickets in our name. So, if you can code or build stuff, grab a place now and come join the fun:

Here are a few links of interest:

O’Reilly prescient on the DataPortability issue

I just found this nugget while scanning for a quote to use from Tim O’Reilly’s 2005 article “What is Web 2.0“:

“Much as the rise of proprietary software led to the Free Software movement, we expect the rise of proprietary databases to result in a Free Data movement within the next decade. One can see early signs of this countervailing trend in open data projects such as Wikipedia, the Creative Commons, and in software projects like Greasemonkey, which allow users to take control of how data is displayed on their computer.”

I’m not sure he imagined it happening in the social data game at the time he wrote this, but it’s pretty clear he was grokking the issues at both macro and micro levels. He at least understood the dynamics that led us into today’s DataPortability debate.

Local community data reporting

EveryBlock has taken a very data intensive look at local news reporting. As founder Adrain Holovaty explains:

“An overall goal of EveryBlock is to point you to news near your block. We’ve been working hard to do a good job of this so far by accumulating public records, cataloging newspaper stories and pulling together various other geographic information from the Web.”

This generally takes the form of raw data points placed on maps. They recently rolled out a variation on the theme by using topic-specific data which adds more context to the local news reporting idea.

“A week or so ago, 15 people were arrested on bribery charges as part of a federal probe into corruption in Chicago city government. We’ve analyzed U.S. Attorney Patrick J. Fitzgerald’s complaint documents and cataloged the specific addresses mentioned within. On the project’s front page, you can view every location we found, along with a relevant excerpt from the complaint. You can sort this data in various ways, including a list and map of all the alleged bribe locations.”

This is the type of value that’s otherwise kind of missing from the experience. Rather than providing a mostly pure research tool, the site now gives some insight and perspective with an editorial view on the data. In this case, the data is telling a story that otherwise might seem a little distant to you until you see how the issue may in fact be a very real one right in your backyard, so to speak.

But it occurred to me that the community is probably even better able to capture and share this level of useful insight. It would be really neat to see EveryBlock open the reporting and mapping process so that anyone who has an interest in exposing the trends in their neighborhood or elsewhere had a platform to do so.

Average payment (€) by Area
Similar to the way Swivel allows you to collect data in spreadsheet form, visualize it and then share it the way Flickr and YouTube allow you to share, EveryBlock could provide an environment for individuals to do the reporting in their neighborhood that matters to them. The wider community could then benefit from the work of a few, and suddenly you have a really powerful local news vehicle.

This isn’t necessarily in contrast to the approach has taken by aggregating shared information from around the web, but it certainly puts some structure around it in a way that may be necessary.

Managing a community is a very different problem than aggregating and presenting useful local data. But I wonder if it’s a necessary next step to get both of these fledgling but very forward-thinking local media services closer to critical mass.

Openness, evil and reusability

I’ve stopped blogging over the last several weeks as I uprooted my family and moved to London to start my new job. But there have been some interesting things worth tracking recently I thought I might mention.

(Interestingly, Twitter usurped any blogging impulses I’ve had during the transition, but it’s time to get back into the long form dialog a little again now that we’re settling in here.)

First, I’m really pleased to see Yahoo!’s open strategy taking shape with things like SearchMonkey, Glue, and the forward-looking presentations done at Web 2.0 Expo. In my opinion, they are still underestimating the power of what Yahoo! could be doing by opening outwardly more, but the momentum is definitely in the right direction regardless of the distracting M&A discussions.

Second, I love where Umair Haque is going with his ‘Good vs Evil’ strategic thinking stuff. He’s getting into why the costs of evil are starting to outweigh the benefits in a globally networked and highly elastic economic landscape.

“As Starbucks and Wal-Mart are discovering, orthodox strategy was built for an industrial world – an equilibrium world of oligopolies, soulless “product”, and zombified “consumers”. But that’s not today’s world.”

Even better than his post, perhaps, is the comment stream which includes this insight from Mike Bonifer who compares today’s competitive landscape to the art of improvisation:

“What many do-gooders fail to acknowledge is that it is not enough to do good. One must also confront, then work artfully at marginalizing, out-witting, out-designing and out-performing the forces of evil that are afoot in the world. Forces like greed, hate, terror, racism, misunderstanding, obfuscation and fear. Heroism is only as strong as the calumny it overcomes.”

Third, I loved hearing the meaty thinking going on in the heads of Lucas Gonze and Jon Udell talking on IT Conversations. It’s as if they are both articulating Clay Shirky’s cognitive surplus view of the world through a music lens:

“Imagine that we lived in a world where all photography was the kind you see in magazines. In this world all photos are taken by professionals and all the people who got their pictures taken are models at the peak of their career. If you had your picture taken normally, you’d think you were hideously ugly. That is the musical world we grew up in, and it’s bogus. Things don’t have to be that way.”

Jon naturally moved the conversation to the problem of discoverability that has been increasingly difficult to deal with as more and more data builds out across the network. He notes some of the challenges as a consumer of interesting things and as someone who has something interesting to offer. He thinks the answer is a bit higher level than traditional syndication:

“There’s a way of publishing that allows something to flow on the network retaining its full fidelity and usability in other contexts.”

Lastly, the open data services space is getting really really interesting now as context and relevance find their way into the mix. For example, the Dash GPS formally rolled out their open service. And then a Guardian colleague pointed me to the AMEE service (“Avoiding Mass Extinctions Engine”) which finds itself being used on Dopplr and the the current Radiohead tour web site (click on ‘Carbon Calculator’).

There are tons of interesting developments unfolding, and I’m seeing all this stuff through fresh eyes again…one of the great benefits of changing jobs. I’ll do my best to keep the blogging energy up and to provide some analysis. Though I’m sure my perspective will shift a bit…to what, I really don’t know, yet.

Good presentation from Yahoo! CTO

I was really pleased to see Yahoo! make a public statement about some of the work going on internally to transform the way the company operates and the resulting changes in the product line. As Cody Simms added in his blog post on YDN:

“We’re moving from a model in which each Yahoo! property develops much of its own technology to one where we share common data and frameworks that can be easily surfaced across multiple Yahoo! properties and off the network. It’s a major rewiring of Yahoo!. “

Here’s the video of Yahoo! CTO Ari Balogh’s keynote at Web 2.0 Expo:

Now, there wasn’t a lot of time in Ari’s presentation. That meant that he had to focus and left out some important pieces, in my mind.

First, he definitely understated the cluefulness the company has shown with things like Fire Eagle, MyBlogLog, Y! Live, OpenID and Hadoop support, Pipes, YUI and, of course, Flickr, to name a few. Yahoo! has innovated around openness and Internet services for a long time. The company needs to be proud of those things and recognize that the advances on these small teams really matter. In fact, it’s those types of advances that I find most compelling in Yahoo!’s arsenal as opposed to the things that help me “get distribution on Yahoo!’s monumentally popular properties”.

I was also a little disappointed that Ari didn’t at least touch on how the company intends to serve the wider Internet and all the activity happening outside of There are a lot of ways everyone benefits from Yahoo!’s data, tools and services wherever those things are used, but he focused exclusively on what happens on

Lastly, it would have been nice to hear about how developers can benefit from Yahoo!’s ad platform or to hear whatever other ways they can build businesses in the Yahoo! ecosystem. All things considered, I think that was a forgivable omission. I hope they don’t take too long to answer that question, though.

My criticisms here are purely driven by my desire to see Yahoo! win. Again, I know there’s some great stuff happening there, and I love seeing the company stand up and speak with confidence about its future. It should do more of that.

I’m looking forward to seeing all this unfold in the coming months. Well done, Yahoo!.

Step 1: Define vision. done
Step 2: Make it so.

Interesting perspectives from Web 2.0 Expo

Today’s Web 2.0 Expo in San Francisco provided some really good brain food.

Clay Shirky’s keynote was excellent. He talked about architecting a new world for the “cognitive surplus” that’s emerging as people pull themselves out of the historical sitcom hangover and invest their energy online. Matt Jones and Tom Coates shared some neat ideas on design for personal infomatics. And Twitter’s Alex Payne and Michael Migurski of Stamen Design presented learnings from the perspective of an API provider.

One little nugget I really liked was a minor point Migurski made when talking through the Oakland Crimespotting service. He noted that there are several standard formats commonly provided by most web services including HTML, JSON, serliazed PHP, RSS and XML.

But we often forget about simple Excel spreadsheets.

He showed how the Oakland Crimespotting site offers downloadable Excel spreadsheets detailing recent activity from particular police beats, for example.

One of the keys to opening up government data is making the case to the people who are best equipped to provide raw data that it needs to be posted directly to the Internet. Telling them they need to output JSON for data visualizations and mashups will do as much good as a slap in the face. Showing them a regularly updating Excel spreadsheet that is findable on a web page that they can email to their colleagues, friends and families is going to get them thinking differently and perhaps encourage their participation directly.

The crime data issue is going to be a big deal in the not too distant future, I’m sure. And as Mr. Coates and Mr. Jones noted in their talks on personal data design, it’s the details that really matter in this space. You can think about products and features all day, but the specifics that define how data is shared, how it becomes relevant and how it is presented will make or break the intent of any offering.

Designing Your API, Web 2.0 Expo 2008: