capitalBLOG

Derivative Networks: Hot Board Member, Angel, and VC Relations

Dean Krikorian — Wed, 11 Apr 2007 02:31:18 +0000

Continuing from the previous post…We now use the relations found in the previous map to investigate ‘hot company’ networks in new ways. This is based on the idea that key people and investors can help point the way to new companies. The goal of a ‘derivative network’ is to take a network of interest and ‘invert’ it to see another view of its affiliations. In each of the ensuing three derivative maps, we will be exposed to new sets of people and companies that are not in this group of ‘hot companies,’ but are related to that group through these derivative networks.

First let’s look at the networks of the three key board members identified in the previous post: Reid Hoffman, David Sze, and Bill Gurley. Q: What do these board members have in common? The goal here is to identify companies not found in the previous post map - companies that are found in the same circles as the ‘hot companies’ discussed in the last post. Along with the new people we find, all of these companies are potentially interesting, so let’s see what we can dig up.

In the last post we started with a map of company nodes. In this first derivative map we will look at the network formed by the affiliations of those companies - namely, the key board members.

In the map above we see the combined network for our three key board members: Bill Gurley, David Sze, and Reid Hoffman. Some key features to note are the central position of Reid Hoffman as an investor which mirrors his importance as a board member in this network formed around the ‘hottest companies.’ Most of the companies from the previous post appear as well (as is expected). Some new names also appear - Marc Andreessen, Ron Conway, Peter Thiel, and Gil Penchina as angel investors; John Dean, Jed Smith, Keith Rabois, and Brett Bullington as board members, and PayPal as a spawning ground for all of this activity. Some of the new companies we are exposed to include: Softcoin, Marquee, Wink, Vendio Services, Six Apart, Open Table, and Tropos Networks. The next map shows the network for the key angel investors from the map in the previous post.

The second derivative (well, kind of). In the map above we’ve mapped the four key individual investors from the previous post. Just as the first map in this post shows a derivative network of the key board members, this map shows the derivative network of the key angel investors. Looking at this map we see that for our network of the ‘hottest companies’ Bill Gurley and Mike Maples are less integrated. Mike Maples isn’t seen in the board member based map above, and Bill Gurley isn’t seen in the investor based map above, but all of the other key players are present in both maps, as well as Peter Thiel, Keith Rabois, and Marc Andreessen. The other point of interest lies in the new companies introduced, especially those with multiple connections such as the Facebook-Powerset cluster between Peter Thiel and Reid Hoffman. Their common involvement in Facebook negates the interest in Friendster to a great degree, but Powerset is an interesting new inclusion. Similarly, Xoom, Revision3, and Wikia are new companies introduced on this map. The next map shows the network for the key VCs from the map in the previous post.

The third derivative (kind of). In the map above we see the co-investments of the four key VC firms identified in the previous post. Along with the ‘hot companies,’ and other companies of interest that we identified from the second map in this post, here we see more companies of interest due to the co-investment relations between our key VCs. It is interesting to note that none of these companies have received funding from more than two of these VCs, which perhaps would have been an indication of a higher level interest or importance in the network of ‘hot companies.’ From this point we could continue to investigate the networks of these other companies - Registered users can do so by clicking through on our maps.There is a tendency in the Valley to fixate on the hottest new thing, and the biggest payoff, but the derivatives of ‘the big ones’ can often lead to a number of extremely successful companies. In any case, the goal of this post was to illustrate alternative viewpoints using network maps (i.e., from the VC, Angel Investor, or Board Member perspective). In this case, derivative networks were used to identify hot company relations. The main premise is that it is not only what you do, but who you are associated with, that helps determine company success. This can be applied to any set of people, companies, or investors in your own area of interest.

Week in Review: Zazzle Zooms, NeoPath + Cisco, and VC Metrics

Dean Krikorian — Sat, 17 Mar 2007 13:52:59 +0000

Zazzle zooms in Alexa Rankings: The Top Alexa Mover and Shaker for the week moved up from 4,028 to 760 (granted, Alexa is only one measure). Zazzle personalizes t-shirts, coffee mugs, stamps, etc. The movement is in no doubt due to a big ad campaign, but they are nevertheless generating some pageview interest. Here is Zazzle’s Relational map:

Zazzle Relational Map. Notes: (1) Zazzle is funded by KPCB and Sherpalo (their only 2 investors); (2) John Doerr and Ram Shriram sit on the board (their only 2 outside board members); (3) Zazzle has capital relations with Google, Tellme, and Podshow (it is a full Kleiner-Sherpalo sibling); and (4) Zazzle and Google are the only two companies that have both Doerr and Shriram on their current board.

Cisco buys NeoPath, a network storage company (Venture Beat)

NeoPath Relational Map (Pendants dropped, Inactive companies included). Notes: (1) Cisco was an investor in NeoPath; (2) Cisco relational investment ties to August Capital = BroadLogic & Ubicom; (3) All of the active companies are private; (4) Only 1 inactive acquired company = Caw Networks; and (5) Cisco relational investment tie to Bill Joy (Caw).

Another qualitative VC Ratings service launches

VC Ratings @ thedeal.com reports on the company The Funded, which offers ratings for VC Firms. A blurb on the post that drew my attention:

While the news and opinion in this blog hopefully provide readers with a sense of which VCs are doing well (and badly) and why, there is little quantitative analysis that would provide statistical-backed insight into not just which firms are doing good work, but which individual partners adds value to those around them.

I believe that our Relational Capital Index would be a good metric along these lines. The main idea here is the quantification of network relations based on the strategic value of VC firm connections. More to come: (1) we plan on publishing our rankings of Silicon Valley Investment and People (Board) Relational Capital; (2) we will incorporate network metrics as an optional advanced feature on LinkSViewer. For those who want to learn more, please contact us at linksviewer@groupscope.com.

The Small World of March Madness: Network Predictions

Dean Krikorian — Thu, 15 Mar 2007 23:11:18 +0000

It’s NCAA Basketball Tourney time - get your brackets ready and start talking some trash. I did a network analysis of the tournament teams and my predictions are below (explanation follows).

Final Game: Florida over Washington State in a squeaker.

Final Four: Florida, Washington State, Ohio State, Pittsburgh

Teams that could go further than expected: Oregon, Louisville, Virginia, Boston College

Teams that could leave sooner than expected: Kansas, Memphis, Texas

Potential First Round Upsets: Long Beach State over Tennessee, Georgia Tech over UNLV

So how did I come up with these predictions? With network analysis, of course (what else do I do in my spare time?). Now, it is common knowledge that who-plays-whom is often used in determining sports rankings (e.g., a strength of schedule indicator). My main premise is that given relations between a finite set of events, network analysis provides a powerful strength of schedule indicator. This is a new way to model prediction markets and the main implication for this blog is that such a model can be applied to financial markets as well (using relational capital as the basis).

Granted, every model needs a theoretical basis and for this I turn to the research literature on Small World networks. Small Worlds have gained in recent popularity based on the original sociological research of Stanley Milgram and modeled in the physics realm, most notably by Duncan Watts and Steven Strogatz. They argue that the small world phenomenon occurs in two network conditions: (1) high local clustering; and (2) small global path lengths. Small world networks are visualized as pockets of nodes grouped together with relatively few links required to reach all others (on average). My goal was to statistically model small world networks.

Back to basketball…If we apply Watts and Strogatz’ finding to March Madness, this means that those teams that have relatively high local clustering and short global path lengths are best prepared for a round-robin (small world) tournament. This implies that teams that form relatively isolated clusters with a few strategic links to well-connected others have a distinct advantage.

Let’s look at a few examples in my NCAA 2007 network graph. Gonzaga has played a number of tournament teams, but they form no distinct local clusters. The PacTen has a small world network advantage because of a local cluster with strategically positioned ties to key gatekeeping teams (they are tied together and connected to, but apart from, the central core). This is particularly advantageous for Washington State and Oregon, who both played Gonzaga. Long Beach State has played both USC and UCLA (high clustering as connected to the PacTen), but have a longer path length because these are their only two connections in the field.

In my model, I also factored in some other stuff. You can check out my resulting bracket here. We’ll see what happens (and good luck to all).

Omidyar Network Revisted (Corrections to the Wikia Board) + Wikia & SocialText Ties

Dean Krikorian — Tue, 13 Mar 2007 22:45:32 +0000

First, a mea culpa on some incorrect information related to the Wikia Board of Directors on our previous post on The Omidyar Network Network. This was pointed out by Ross Mayfield (CEO of SocialText) and also commented on by Marc Canter on his blog - if anyone would know it was wrong, then it would be Ross (he’s on the map!). Our information from the LinkSV database has since been updated (sorry for the inconvenience). The changes result in only one direct connection from Wikia to SocialText (Jimmy Wales, Wikia Founder) - no cluster of common advisory board members (as noted before), but there still exists a tie. The corrected map is shown below (and also updated on the original Omidyar post).

Corrected Relational Network Map (Pendants or one-tie nodes dropped). Note how Jimmy Wales is now the only board connection between Wikia and SocialText. The Wikia board has been updated. The Wikia/SocialText ties are interesting in their own right…

Relational Map for Wikia + SocialText (Pendants included). Notes: (1) inclusion of all board members and management team members for both companies; (2) gatekeeping role of Jimmy Wales and Omidyar; (3) addition of Jeremy Levine (Bessemer) and his board ties to Yelp and Gracenote; (4) additional connecting presence of angel investors Brett Bullington, Ron Conway, Dan Gillmor, Michael Tanne, and Dave Whorton; and (5) hybrid role of Joi Ito as angel investor (Wikia) and board member (SocialText).

What I find most striking from the above map are the extensive ties between the companies in the center of the map (despite both having relatively few VC investors). Certainly there are some alliances that have or will be formed by these related companies - look at the names, a virtual Who’s Who of Web 2.0. Indeed, I commented in an earlier post on the relational capital of Web 2.0 companies and how individual investors, well-connected in the space, have helped redefine the venture capital (and social media) landscape. True to the spirit of blogging, individuals can make a difference (but it helps to have some connected VCs as well).

Miasole, SunPower, Nanosolar cash in on Bush’s solar agenda

Brad Cohen — Sat, 10 Mar 2007 00:55:26 +0000

VentureBeat reported on the Silicon Valley windfall (haha) of cash for solar energy initiatives. Here’s a map for the combined networks of three of the recent grant recipients. Looks like a good spread of involvement across the valley without too much overlap - investors aren’t putting money into redundant solar energy projects. Click on the map below too see a live LinkSViewer version that’s easier to read. Find out who’s benefiting from this windfall of grant money, and who’s related to those beneficiaries.

LinkSViewer 1.2 New Feature: LinkSV People Profile Pop-Ups

Dean Krikorian — Mon, 19 Feb 2007 03:16:26 +0000

We recently updated LinkSViewer to include LinkSV people profile information. The people profiles, company summaries, and investment firm summaries are also now all available from the Information Window in LinkSViewer (which we spruced up a bit). This should improve the functionality of the site and speed up the navigation process for users.

By double-clicking on any person node in the map, their LinkSV.com profile pops up in a new window. The profile allows immediate access LinkSV’s Web-accessible information on their background, boards, and investments. It also links to other available information such as their Web bio, company Website + address, LinkedIn record, and Jigsaw contact information. Here is an Example LinkSV Profile (full-screen).

Double click on a person (board or management team member) the next time you map, it should be useful. More new stuff to come…

Midas List Network Analysis Summary: Lessons Learned and Overall Trends

Dean Krikorian — Wed, 31 Jan 2007 02:08:24 +0000

This post is divided into two sections, Lessons Learned and Overall Trends. The first section contextualizes the network analyses using three questions: (1) Why the Silicon Valley?; (2) Why Networking?; and (3) What are the Predominant Types of Maps? The second section provides details on the overall trends from the Midas List Network Analyses in two parts: (1) Midas List 2006 vs. 2007: (a) New to the 2007 List; (b) VC Firm Changes; and (c) Title Category Changes; (2) Midas List Network Content Analysis: (a) Top Midas List Connectors; (b) Top VC Connectors; Top People Connectors; (c) Top Future Stars + Watch List Private Companies; (d) Top Future Star Private Companies; and (e) Top Watch List Private Companies.

Lessons Learned

After completing brief analyses for 97 of the Forbes Top 100 Midas list, there are several lessons learned from my experiences. I would like to share this knowledge with Capital Blog readers. Lessons learned are divided into three sections: (1) Why the Silicon Valley?; (2) Why Networking?; and (3) Types of Maps.

Why the Silicon Valley?

Because the LinkSV database we draw our maps from is mostly regional in nature, it should be noted that the scope of our analysis is limited to Silicon Valley companies. The majority of dealmaking in the high-tech and life sciences centers around the Silicon Valley, although this trend has presumably been shifting in recent years. We do not claim to be the end all analysis for the Midas List, only to note the importance of Silicon Valley firms and their people. Our main premise considers the Silicon Valley as the Venture Capital center of the world and this is indicative in the delivery of 97 relevant maps for the top 100 Midas List members. In the near future we plan to expand our scope to the US and worldwide, along with our partner LinkSV, so stay tuned on this end.

Why Networking?

The Silicon Valley has long been considered as a highly networked arena for Venture Capitalists - Sand Hill Road is said to be the epicenter of the dealmaking world. Because of the proximity of VC firms and their people, high-tech networks tend to be more connected than the typical business world. Thus by mapping such networks, one gains insight on how this connectivity is translated into investment structures by mapping who is connected to whom and by what. At GroupScope, our goal is to make these structures visible to the end user by translating Social Network Analysis research (I was a Social Networks professor at Cornell University) into tangible meaning for the everyday business user (in this case investors, start-up companies, and venture analysts). The main purpose of the network schema presented in the introduction is to provide everyday insight into important capital connections. The purpose here was to cover a broad spectrum of network indicators that highlight the important facets of the network maps:

- Ties: At least 2 connections between investors, companies, and/or people
-  Strong ties: At least 3 connections between investors, companies, and/or people
-  Key paths: Reachability and shortest path lengths
-  Key clusters: Large groupings in the network

There is also a predictive nature in the network analyses. This is based on the main premise that success breeds success - that successful partnerships in the past presumably translate into successful partnerships in the future. Future Stars are Private Company network indicators based on co-investment structures between investors and board interlocks between people. The main idea of a Future Star is contingent on the positioning of a private company in the network and whether it is tied to past successes (Public or Acquired company ties): Blue boxes connected to Green or Purple boxes. Watch List Private Companies have direct ties to the target individual/VC firm, but may not have multiple ties to Public or Acquired companies. The goal heere is to identify Private Active Companies that are tied to the Midas list as indicators for future success, particularly as a guide for late stage investors.

What are the Predominant Types of Maps?

Based on my experiences in choosing and analyzing the network maps for the Midas List, there seems to be a predominance of five general types of maps:

Individual Board Member Maps: Individuals who have multiple board seats on Silicon Valley companies. Consider these maps as Board of Director Interlocks or Investment Firm Interlock (when appropriate).
Company and Individual Maps: Individual board members who are tied to a key company that is mentioned in their Midas List bio. Here, the pathways between the company and person play an important role as well as populating more sparse network maps.
Individual and Investment Firm Maps: Individuals and their ties to their present investment firms. The combined connections play an important role by linking the people to their investment firm because the investment firm may play a more important role for those individuals who do not sit on boards despite playing a large role in brokering deals for their firms.
Investment Firm Maps: Investment firm ties used in the case where individuals do not sit on boards of Silicon Valley companies. This provides a global network map for the investment firm in cases where the individual does not have direct ties to Silicon Valley firms.
Liaison Maps: This is the case for bankers or lawyers in particular, where the ties to their investment firm are usually less important than the deals they broker. The emphasis here is on linking key deal companies together and linking their paths together through key people, companies, and investors (i.e., how to get from Point A to Point B).

Overall Trends

Below are two groups of overall trends: (1) Comparing the Midas Lists of 2007 and 2006; (2) Compiling the categorical trends for all 97 Midas List network analyses.

Midas List 2007 vs. 2006

These comparisons examine basic differences gleaned from the 2007 and 2006 Midas Lists: (1) New to the 2007 List; (2) VC Firm Changes; and (3) Title Category Changes.

New to the 2007 Midas List (#)

Ronald Conway (#7)
Jixun Foo (#11)
Asad Jamal (#12)
Mark Tluszcz (#20)
Roelof Botha (#23)
Philip Nicholls (#26)
Kheng Lee (#29)
William Ford (#30)
Peter Barris (#32)
Gordon Ritter (#34)
Shlomo Kalish (#39)
Allen Miner (#40)
Harry Weller (#46)
J. Sanford Miller (#49)
George Lee (#57)
Navin Chaddha (#58)
Thomas McConnell (#66)
Guy Nohra (#73)
Tench Coxe (#75)
Michael Krupka (#76)
Mike Gordon (#77)
Michael Ross (#81)
Brent Ahrens (#87)
Fred Cohen (#88)
Yossi Sela (#89)
Michael Bronfein (#90)
Augustus Tai (#92)
Jack Lasersohn (#93)
Peter Fenton (#94)
Ravi Mhatre (#95)
Ned Hooper (#96)
Kathleen LaPorte (#97)
Ralph Sabin (#100)

33 People are new to the 2007 Midas List – they were not on the 2006 Midas List. Ron Conway made the largest jump to #7 (with a bullet).

Top VC Firm Changes 2007/2006 (#)

New Enterprise Associates (5/4)
Bessemer Venture Partners (4/4)
Granite Global Ventures (4/3)
Accel Partners (3/3)
Alta Partners (2/1)
Battery Ventures (2/2)
Benchmark Capital (2/2)
Canaan Partners (2/1)
Emergence Capital Partners (2/1)
Goldman Sachs (2/3)
Lightspeed Venture Partners (2/1)
Morgan Stanley (2/2)
Norwest Venture Partners (2/2)
Panorama Capital (2/0)
Sequoia Capital (2/2)
Summit Partners (2/2)
Sutter Hill Ventures (2/2)
Oak Investment Partners (1/3)
Clearstone Venture Partners (1/2)
Doll Capital Management (1/2)
InterWest Partners (1/2)
Kleiner Perkins Caufield & Byers (1/2)
Menlo Ventures (1/2)
Highland Capital Partners (0/2)
JP Morgan Partners (0/2)
Redpoint Ventures (0/2)
Worldview Technology Partners (0/2)

The Venture Capital Firms (Companies) at the top of this list were the most represented on the 2007 Midas List. The first number in the parenthesis is the 2007 representatives and the second number is the 2006 repreresentatives. Note the 4 firms at the bottom of the list that had two representatives in 2006 and zero in 2007 (Highland Capital Partners, JP Morgan Partners, Redpoint Ventures, Worldview Technology Partners).

Top Title Category Changes 2007/2006 (#)

VC = Venture Capitalist (82/80)
A = Angel Investor (7/5)
B = Banker (4/5)
BU = Buyout (4/6)
L = Lawyer (3/2)

These titles were the most represented on the 2007 list. Note the slight increase in Angel Investors and slight decrease in Buyouts from 2006 to 2007. For the most part, the category number remained about the same.

Network Analysis Trends/Content Analysis

This section summarizes the Network Analysis Trends for the 97 Midas list network maps. Her I compiled a content analysis of all of the network maps. It should be noted that the types of maps affect the trends, but that the content reflects connection representation. These lists indicate the most connected on the maps based on their direct and indirect ties to other members/firms from the Midas List. Total mentions indicate the number of appearances on all network maps under the specified category.

Midas List Top Connectors (Total Mentions)

David Cowan (5)
John Doerr (3)
Arthur Patterson (3)
Ram Shriram (3)
Ron Conway (2)
Michael Moritz (2)
Lip-Bu Tan (2)

These Midas List members were most connected to other Midas List members. The number in parentheses indicates how many times they were mentioned as Midas List Connections in all network analyses. It makes sense that the Top Midas List members are well-connected to other Midas list members. Note the importance of David Cowan based on his role as Bessemer Capital’s Silicon Valley representative. Four out of the six top Midas List members (Doerr, Shriram, Conway, & Moritz) are among the top 7 connectors. The Google tie makes sense as a common thread between top Midas List members, but each person must have at least one other connection to make the above list.

Top VC Connectors (Total Mentions)

Sequoia Capital (19)
Kleiner, Perkins, Caufield, & Byers (15)
Accel Partners (13)
Benchmark Capital (9)
Hambrecht & Quist (9)
NEA (9)
Intel Capital (8)
U.S. Venture Capital (7)
Venrock Associates (6)
Lightspeed Ventures (6)
Goldman Sachs (5)
Mohr, Davidow Ventures (5)
Chase Capital (4)
Focus Ventures (4)
InterWest Ventures (4)
Mayfield Fund (4)
Mobius Ventures (4)

These investment firms were most connected to other Midas List members. The number in parentheses indicates how many times they were mentioned as VC Connections in all network analyses. Note the dominance of Sequoia, Kleiner, and Accel as those VC firms with a large number of portfolio companies. This reflects the importance of their Relational Capital – that their connections are forged with key people and VC firms, as evidenced by their connections to those on the Midas List.

Top People Connectors (Total Mentions)

Bandel Carrano (3)
Nancy Schoendorf (3)
Greg Avis (2)
Allon Bloch (2)
Bill Campbell (2)
Rob Chandra (2)
Peter Chung (2)
Bob Corey (2)
Peter Currie (2)
John Grosshans (2)
Thomas Kiley (2)
Sanjeev Malhotra (2)
David McCallum (2)
Arnold Oronsky (2)
John Ryan (2)
Lowell Sears (2)
Kip Sheeline (2)
Casimir Skrypczak (2)
Eckard Weber (2)

These people did not make the Midas List, but are connected to those that did make it. The number in parentheses indicates how many times they were mentioned as People Connections in all network analyses. These are people to keep an eye on for future Midas lists. Keep an eye on Bandel Carrano and Nancy Schoendorf.

Top Future Stars + Watch List Connections (Total Private Company Mentions)

SuccessFactors (5)
*2Wire (4)
*Friendster (4)
Plaxo (4)
Postini (4)
*Rapt (4)
*Visto (4)
AGEIA (3)
AtHoc (3)
Beatnik (3)
*Digital Chocolate (3)
Mellanox Technologies (3)
*Netli (3)
Next Medium (3)
Nominum (3)
Novacea (3)
Novasys Medical (3)
Omneon Video (3)
OpenTable (3)
TransOral Pharmaceuticals (3)
*Veraz Networks (3)
Yipes Enterprise Services (3)

* indicates companies on Always On’s most recent Private Company or Media Top 100.

These private companies were most connected to those on the Midas List. This is an indicator of high Relational Capital for these companies because it means they are tied to key individuals and investors. Future Star companies are tied to Public or Acquired companies that have at least two connections to people or investors from the Midas List. Watch list companies have direct ties in Midas List network maps, but may not have two or more ties and/or links to Public or Acquired companies. It is interesting that many of these companies were honored by Always On, especially since many of these companies are Medical and Biotech companies not under consideration. This list provides unique network indicators of the strategic alliances that can guide private companies toward fruition. It can be especially be used by investors considering late rounds of financing or secondary directs. Keep an eye on the top eight on this list (SuccessFactors, 2Wire, Friendster, Mellanox, Plaxo, Postini, Rapt, & Visto).

Top Future Stars (Total Private Company Mentions)

*2Wire (3)
Beatnik (3)
Mellanox Technologies (3)
*Veraz Networks (3)
*Visto (3)

* indicates companies on Always On’s most recent Private Company or Media Top 100.

This is a breakdown of the aforementioned list emphasizing multiple connections to Public or Acquired companies in Midas List network maps. These companies are evidence of prior success (IPO or exit) for Midas List members and/or their firms – they have multiple ties leveraging past success. The supposition here is that the Midas Touch is experiential. Keep close eye on Veraz Networks, which recently filed for IPO status.

Top Watch List Connections (Total Private Company Mentions)

Postini (4)
AtHoc (3)
*Netli (3)
NextMedium (3)
Novasys Medical (3)
*Rapt (3)
SuccessFactors (3)

* indicates companies on Always On’s most recent Private Company or Media Top 100.

This is a breakdown of the aforementioned list emphasizing direct ties in Midas List network maps. These companies have direct ties in Midas List members and/or their firms, but may not have two or more ties and/or links to Public or Acquired companies - they have Midas ties. The supposition here is that the Midas Touch is contagious. Keep a watch on Postini.

This concludes the series on the 2007 Midas List networks. Feedback appreciated - Feel free to drop us a note on what portions of the analysis were most useful. As noted in the introduction, if you have any further questions or want more detailed Midas List network analysis reports (including commentary and/or statistics), please email us.

JotSpot hits the spot for Google

Brad Cohen — Tue, 31 Oct 2006 22:57:57 +0000

Google’s acquisition of JotSpot is a great little trick or treat surprise for Joe Kraus. Tons of blogs out there are already covering their predictions of what the incorporation of this wiki platform will bring to the already significant suite that Google has been piecing together… spreadsheets, word processing, online collaboration, and more - all under the umbrella of Google with their uncanny knack for optimal monetization. But that’s only moderately interesting to me. If I got incredibly interested and surprised every time Google made a move that sounded potentially groundbreaking and innovative, I wouldn’t have time to do anything else. The networks involved, however, do interest me.

What I see when I map Google and JotSpot together is the fact that they’re not exactly connected in any direct and profound ways. Let’s add a couple of caveats in there… Joe Krauss (formerly Excite.com as well as many other ventures of interest) is an extremely well connected individual in Silicon Valley, so it is doubtless that there are links that exist beyond what can be seen in these network images. Let’s take a look.

Relational Map of Google, JotSpot, Infrasearch, and Joe Kraus
(click to see full size)

If you click on the image above and examine the full size map you will see that Google and JotSpot are not all that closely related (at least in terms of common investors, common board or management team members). JotSpots shortest pathways to Google are still 4 degrees of separation - that is, Google to an investor to an investment to another investor or board member to JotSpot.

Now look at the top right of the map and you will see the yellow node that represents Joe Kraus as an investor. What’s interesting here is that his investment network is so disparate from the network around JotSpot. This definitely works to the advantage of Joe as he can exploit two completely non-redundant networks whenever he’s trying to get to anywhere.

Infrasearch is also included in the map. It is an inactive acquired company that connects Joe to Angel Investors, LLP, and then to Google.

It is abundantly clear that Joe and JotSpot have connections to many powerhouse companies and investors like LinkedIn, Marc Andreesen, Mayfield, and Redpoint, so I am not trying to argue that JotSpot is isolated and we should all be surprised by the fact that Google gathered them into the fold. However, it is not a ‘no-brainer’ just by looking at the networks here. It tells me that this acquisition is important to Google, potentially moreso than some others. If Google takes the time to wander out of their direct circles to swallow up JotSpot, then this wiki platform might really be something more than just a face value wiki functionality. And perhaps all those bloggers going nuts about how this wiki platform is going pull everything together and let Google compete with everyone from Microsoft to Salesforce.com.

As an aside, here’s a great link from a panel moderated by Guy Kawasaki. It includes Reid Hoffman, the guy from JahJah, the guy from Photobucket, and Joe Kraus. It’s all about the startup adventure, and there are a few great nuggets of wisdom from Joe about persistence in the face of adversity.

Link to Guy’s blog to see the video

Boo! Happy Halloween.

Pals from PayPal, Payoffs, and Playoffs

Brad Cohen — Thu, 19 Oct 2006 06:11:12 +0000

I have to mention that the Mets just won to extend the series to it’s final game. Not that I’m such a huge baseball fan, nor a Mets fan, but including the word “Playoffs” in the title allowed it to roll off the tongue. On to the networking…

Miguel Helft, who used to be often read in the Mercury, is now a tech writer for the NY Times and had an article on the front page business section on Tuesday about the pals from PayPal and their microcosmic Silicon Valley networking effect, and how that small group has hatched and reared some of the most successful companies (and exits) in the tech world. However, Helft is not yet familiar with our LinkSViewer tool that does visual networking for Silicon Valley tech and VC networks. As such, I can’t blame him for the illustration from the times - a network of sorts, but I can credit him with the inspiration to put up some of my own constructed networks of this power cluster that has achieved the kind of success to which we all aspire.

Here is the first network view. It shows the network formed by including Peter Thiel as an investor, Reid Hoffman as an Investor, and PayPal as a company. Simply including these three nodes creates a network that includes most everyone mentioned in Helft’s article.

Relational Map for Thiel, Hoffman, and PayPal (click to see full size map)

It is important to remember that in this network both Peter Thiel and Reid Hoffman are represented twice; once as an investor with a yellow node, and once as a person (board or management team member) by a pink or purple node - as noted in the legend at the bottom of the full size map. In the center of the map, the three person nodes for Thiel, Hoffman, and Roelof are very central because of their involvement with companies across this entire network. The investor nodes for Hoffman and Thiel create the bottom section of the network because of their investments in companies like Wink, Facebook, SixApart, and LinkedIn.

In the following image, the investors are filtered out to clear up the person-company connections. LinkedIn’s node has also been expanded in this view.

Relational Map for Thiel, Hoffman, PayPal and LinkedIn - no investors
(click to see full size map)

In the full size version of this map you can see a pretty clear distinction between the old and the new. With the investors filtered out, the person-company links are clear and easy to see. The bottom left of the image is dominated by PayPal and all of the people linked to it, but the center and top right portions of the map are populated by all of the newer adventures of the PayPal pals, including LinkedIn, Xoom, Facebook, Tagged, Wink, Friendster, and Grassroots Enterprise.

The next image focuses on just the newer adventures after PayPal. This network is constructed around the company LinkedIn; the people Roelof Botha, Peter Thiel, and Reid Hoffman; and the investors Reid Hoffman and Peter Thiel.

Relational Map for Thiel & Hoffman as people and investors, Botha as a person, and LinkedIn as a company (click to see full size map)

In this network you can see the powerhouse of company involvement around these individuals. Hoffman, Thiel and Botha are especially important because these networks only include board members and management team. Even though YouTube’s founders are no doubt going to assume more and more central roles in this network if they continue their involvement with new tech adventures, but they were not management team level employees at PayPal.

Here’s one last image which includes all of the pendants which were removed from the previous maps. This makes the network messier and a little bit harder to sift through for beginners, but it includes more information, including nodes for Steve Chen, Chad Hurley - two of the three co-founders for YouTube (Jawed Karim will be included the next time the data set gets updated, but I didn’t want to wait to post this).

Relational Map for PayPal, Hurley and Chen as person nodes, and Hoffman and Thiel as investors - Pendants Included (click to see full size map)

Not only can we now see the connections for Chad Hurley and Steven Chen, but Sequoia Capital’s key involvement in many of these companies by way of Roelof Botha is also clear.

Ok, so I haven’t settled on one perfect representative network, but I stand by my claim that there is more information in these. What do you see in these networks?

ADDENDUM:
Dean was playing around and provided a few other interesting network maps that he constructed in LinkSViewer by including the nodes for PayPal, Hoffman and Thiel as investor nodes; and Hoffman, Thiel, Botha, and Rabois as person nodes. Here are the links if you’d like to check them out as well.

Dean’s Maps
Map 1 - Company Map (Filter = Post-2001 start companies)
Map 2 - Relational Map (Filter = Post-2001 start companies)
Map 3 - Person-Investor Map (Filter = Post-2001 start companies)

HP Board Controversy, Board Isolation, and Groupthink

Brad Cohen — Mon, 25 Sep 2006 04:16:56 +0000

Important Disclaimer:
It is important that readers do not automatically assume that just because a person or company appears in this network, it should be somehow affiliated with the current HP board scandal. By nature of how our network analysis works, we are investigating the power of indirect links. As such, it would be rash and irresponsible to assume that a company or person with an indirect link to HP – or in this case, to one or two of HP’s former Board Members, indicates some sort of impropriety in the totality of their work and dealings. The ‘real-world’ nature and context of these relationships must be taken into account. This map can illustrate how those connections are structured, but conclusions as to the nature of those relationships is inference only. It is my responsibility as a conscientious analyst to state this very clearly because I have a concern about how people may jump to conclusions. Viewing networks like this is very new for some people, and with the hype surrounding the connections displayed here we must be sure that when this network is viewed, it is viewed as an illustration of the network structure for HP’s current and former board, and NOT as a direct illustration for a particular incident.

Relational Map for HP’s Board of Directors Network
This LinkSViewer map shows HP’s network, including current and former Board Member connections, and active and inactive companies included. Note that this only includes Northern California companies. The Person nodes for Robert Knowling, Sam Ginn, Tom Perkins, and Robert Wayman are expanded. They are the only four current or former Board Members that link to other companies as well as HP. This means that even if all of the other Person nodes were expanded, no additional links would appear.

What does this map tell us?
1) Of seven former board members, only three (Knowling, Ginn, and Perkins) have other connections as board members in the Silicon Valley. This is a relatively low percentage and indicates a high probability that former HP board members move on to projects and positions outside of the Silicon Valley. Apart from those three, only the CFO (Robert Wayman) has any links to other Silicon Valley companies, and if those three former board members and Wayman are removed from the map, then HP is completely isolated with no connections to anything else in the valley. So any issue that arises is internally magnified because of the lack of correspondence as no channels exist for board knowledge to move to the outside world.
2) Although the portion of the network expanded through Knowling, Ginn, Perkins, and Wayman, does show some interesting overlaps and relationships, there is nothing particularly noteworthy in terms of patterns of connections. Furthermore, because three are former board members, it is even less likely that anything in that portion of the network has any real connection to the current situation on the HP board. Wayman’s board connections also dead end with the former board members links removed. So the effect is an isolation of the board as a result of insulation from outside ties. This is interesting because the nature of the insulation prevents communication to the outside by any normal channel. Thus, the only way for information to seep out is through a leak – perceived or otherwise.

The nature of the current controversy was caused by secrecy and subterfuge. This indicates that it may have more to do with internal dynamics of the board rather than the connections of the board to other external influences. If we were to remove the former board members from this map (including Knowling, Ginn, and Perkins), is almost a complete isolation from any contact to other boards within the Silicon Valley. The relative isolation of board members of such a large company from other connections in the valley could be attributed to the fact that HP is such a large company that taps corporate resources nationally and internationally. And this isolation makes HP susceptible to Groupthink.

Groupthink is a well-known group theory developed by Irving Janis to explain the faulty decision processes that often occur in small groups (e.g., Bay of Pigs, Watergate). Groupthink may explain contributing factors to the HP debacle (we’ll let the readers be the judges). Below are Janis’ 8 Symptoms of Groupthink:

1. Illusion of invulnerability – creates excessive optimism and encourages taking extreme risks.
2. Self-appointed mindguards – members who protect the group from adverse information that might shatter their shared complacency about the effectiveness and morality of their decisions.
3. Self-censorship – protection against deviations from the apparent group consensus, reflecting each member’s inclination to minimize the importance of their doubts and counterarguments
4. Outgroup stereotyping – enemy leaders are viewed as too evil to warrant genuine attempts to negotiate; or as too weak and stupid to counter whatever risky attempts are made to defeat their purposes
5. Belief in inherent morality - inclining the members to ignore the ethical or moral consequences of their decisions
6. Direct pressure on dissenters – any member who expresses strong arguments against any of the group’s stereotypes, illusions, or commitments is considered as contrary to what is expected from loyal members and pressured to conform
7. Collective rationalization – collective efforts to rationalize in order to discount warnings which might lead the members to reconsider their assumptions before they recommit themselves to their past decisions
8. Shared illusion of unanimity – concerning judgments conforming to the majority view (partly resulting from self-censorship of deviations, augmented by the false assumption that silence means consent)
Source: Janice, I. L. (1972). Victims of Groupthink. Boston, MA: Houghton Mifflin (pp. 197-8).

Researchers would claim that overly isolated and cohesive groups tend to have Groupthink. What happens, however, when the protective insulating layer around the group breaks down? At this point, the group is no longer cohesive and can implode upon its shroud of secrecy.

We invite opinions of others regarding whether this example typifies a Groupthink occurrence or not, and why. What do you think? Do you see other prominent examples of Groupthink happening today?