Occam's Razor

VC selection from the CEO perspective

Several good posting this week on How to choose a VC from a CEO perspective from
Fred Wilson and Jeff Nolan and Feld. and Ed Sim and Matt Blumberg.

To add to the existing dialog, I would also mention to a CEO looking for funding, and wondering how to choose a VC that:

- VC groups have they own lifecycles; a given VC fun has a life of 10 years, and an investment period of 4 to 5 years. It’s important to understand where a group is in it’s cycle, how many years until the end of the investment period, and how many years until the group will go back to it’s investors to raise a new fund. This will have a significant impact on the risk-tolerance of the general partner group, and its patience.

- VC manage they portfolio risk as a whole, a new investment will be a new element in the existing portfolio, and the general partner behaviours (and the pressure of its partner during the regular quarterly portfolio review) will be a function of the overall health of the portfolio. It’s important to due-diligence the overall portfolio, and get a sense of how well is each existing investment doing.

- VC have long memories, in the last 5 years, we have each build a list of A-guys that we trust and value, and will do deals with in the future, and of C-guys that we would not do the next Google with, even if we could. You should get a sense of the general partner relationships, and what open or closes the addition of a new investor to your company.

-Board have a maximum size (in my eyes 5), and before adding a new investor, and a new board member, you should thing about who you will remove to keep your board size at the right level and avoid inflation. I have sat on board of 7 or 9 members, and they are a recipe for trouble, as chance for divergence and dysfunctional behaviours are then too high.

And if/when things will go south; you can always practice the art of firing you board members, according to Jerry Collona...

An Almost on world

Interesting posting from Ed Sim on Enterprise Software Software and Intel views on technology evolution. I totaly agree with such a vision off an Almost-On World were...

Asynchronous XML messages are what makes this service-oriented world possible. We are just at the beginning phases, a new architecture is needed and with that comes new and interesting opportunities for software investments. I totally agree here as most of the service-oriented talk from many of the large tech vendors is still a pipe dream and more marketing than fully functioning product. In addition, most enterprises are experimenting with various aspects of the above themes but far from prime-time in terms of deployment...

A world of
1. Software and data delivered as services
2. Hardware as a virtualized resource
3. Autonomic data sources (RFID, tags, smart sensors)
4. Occasionally connected usage
5. Services cross firewalls (security)

Most of this will be also covered next week during the Web 2.0 conference which will investigate the same trends.

One other good additional post along the same line, but from the developper perspective, on what the platform is to deliver in such a almost-on world from the new blog from Adam Bosworth (former Mr XML at Microsoft, them BEA, now Google..).

WiMax confusions

We have been investigating several WiMax deals recently, some trying to build and deploy new WiMax operators, and some working on delivering better inforastructure to WiMax operators.
As we are progressing our due-diligence, we are struggling with the following issues (opinions, facts, and comments welcome on all of them...).

(1) Does the technology work on the long haul
If WiMax is the next Wireless DSL replacement, how does the performance degrade with the distance ?

(2) At what point in time does WiMax started to work for full mobility and roaming, and becomes a potential replacement for 3G technologies

Several good postings recently on this subject

Jeff Nolan :: Making the Case for WiMax

The Intel Technology Journal has an issue focused on WiMax. link courtesy of FierceWireless They expect to have fixed outdoor wireless in first half (H1) of 2005, with fixed indoor wireless equipment in H2 '05, and integrated laptop and portable gear in 2006. It's interesting that they are going after outdoor first, then ramping down the signal strength for indoor installations...

Ron Gruia :: Some Hurdles Will Have to be Cleared Before WiMAX Takes Off

Besides spectrum, another obstacle for WiMAX is the development of the IEEE 802.16e specification for mobile wireless broadband. This spec (part of the larger WiMAX 802.16 standard) is not expected to be ratified until the end of 2005. Other restraints include base station interoperability and higher costs until that issue is solved and economies of scale can be achieved. ..

Ron Gruia :: Intel Developer Forum Update - Part 2

WiMAX, also known as the IEEE 802.16 standard, is a wireless specification that can deliver two-way Internet access at throughputs of up to 75 megabits per second at long range. WiMAX proponents claim that it can transmit data for distances up to 30 miles between broadcast towers, covering areas having more than a mile in radius, all of this at a speed exceeding current DSL and cable broadband capabilities. Intel envisions three WiMAX models: fixed access (alternative to DSL or cable broadband), portable use (within metro zones) and a full mobile system that includes hand-offs as users move between cells (the IEEE 802.16 committee still has some work to do in delivering that mobility management). The Rosedale design does not include the radio element (Intel plans on working with third-parties on that front) and enterprise functionality (i.e. routing or switching details that will be part of a yet to be finalized specification)....

Om Malik :: TI: WiMAX, or WiFUD

check all the comment posted after Om's article, like this one Damn near everything related to WiMAX is a complete fraud. 1) It will never provide a broadband experience at 20-30miles. The physics are obvious. 2) It not even relevant on the backhaul scene due to two developments, a) WSP's have bulk contracts with ILECs and the new rates show great deflation, b) MULTIPLEX ers, allow carriers to extract 2x-4x the normal T1 yield. Net-net the monthly cost for backhaul is only changing modestly, even though volume has exploded. This is just like whats happening on the retail side. Big surprise. So, if WiMAX is a complete fraud, why does it have momentum? Intel is in a colassol(sp?) bind, with the CPU having long ago entered "overshoot" mode, and after they got the ass' kicked in the GPU market, they have to have some kind of motivation for future PC buyers to buy new PCs....

WiFiNetNews :: France to Get Nationwide WiMax

Altitude Telecom, the only owner of a nationwide 3.5 Ghz license in France, plans to use Alvarion gear to build a broadband wireless network in the country: The Alvarion gear is based on WiMax, though not yet certified as WiMax because the certification process hasn’t started yet. Altitude will start out with four counties and move on from there. It’s interesting to note that Altitude plans to use the wireless network to serve small to large businesses. That target market is the same market that broadband wireless operators have traditionally targeted. WiMax, however, has often been touted as a DSL replacement that could be used to serve the residential market. Perhaps Altitude will use the next generation of WiMax gear to target the residential market...

WiFiNetNews :: TI Doubts WiMax

TI, which is not heavily into WiMax like its competitor Intel, says WiMax won't be very effective at bringing broadband to the home: It's true that it's far from certain that WiMax will be anywhere near the success that Intel promises, but most of the reasons TI gives here are pretty weak. Because China hasn't jumped on the bandwagon and because broadband wireless standards have failed in the past doesn't prove that WiMax will fail. If WiMax products have a lower price tag and are more robust than previous attempts at broadband wireless, the technology has a chance of success. WiMax can be far easier to deploy than most wireline technologies and appears to offer a good alternative to wireline especially in developing regions of the world. But there are plenty of reasons that WiMax could fail, which aren't mentioned by the TI executive. In the United States only a few spectrum holders own the licenses that would be ideal for a WiMax deployment and it's not clear that they're interested in the technology. One of those companies, Nextel, has expressed interest in using other proprietary technologies in the spectrum. Other large operators that don't own such prime spectrum are unlikely to want to execute a major deployment in unlicensed frequencies. The TI executive suggests that a portable or mobile version of WiMax might have a better chance of success. While future iterations of WiMax may sound more promising than the initial fixed version, it's very difficult to know today what the market will look like by the time a mobile or portable solution becomes available. By then, other technologies may have leapfrogged WiMax. It's also not clear that WiMax will indeed result in low-cost equipment for operators. Some wireless ISPs have said that vendors they've spoken to have said that the first couple generations of their base stations won't be interoperable with clients from any vendor. That lack of interoperability may not be very attractive for many operators and may prohibit prices from dropping. Ultimately, there are plenty of forces working against WiMax but much of what the TI executive says here sounds to me like sour grapes....

Pulver telecom 100

As a venture capitalist I love list, I got another one today with the Pulver 2004 100, i went though the list, checking who i knew, and who i had never heard about (try to get a sense of the current coverage of our network vs where the innovation was unfolding...)

As 2004 has been the year of VoIP – and the companies named in the pulver 100 have made significant contributions, by helping to spread IP communications worldwide with little or no dependence on traditional telephone network, equipment, or service providers.The value chain characterized by the pulver 100 differs from the vertically integrated telecom model of the last century, focusing more on a networking-industry model that espouses open interfaces, connectivity decoupled from services, and software decoupled from hardware....

The companies named to the 2004 pulver 100 are:

Access Line, Acme Packet, Aculab, Adax, Amtelco, Atreus Systems, Bay Packets, BlueSlice, Bridgeport Networks, Brix Networks, Broadsoft, Broadvoice, CallWave, Centile, Cicero Networks, Cinergy Communications, CITEL Technologies, Commetrex, Convedia, Diamondware, Digilinea, Digium, EADS Telecom, Ecuity Advanced Communications, eStara, Excel Switching, GrandStream, Hexago, Hotsip, i3Micro, Inclarity, Ingate, International Systems Research, Intertex, Intoto, IP Unity, Iptel.org, IPeria, IVR Technologies, Jasomi Networks, jNETx, Kagoor, locustworld, LongBoard, M5Networks, MailVision, Mediatrix, Mera Systems, Meru Networks, Metaswitch, Natural Convergence, Netcentrex, Netintact, Netrake, Neustar, NexTone Communications, Nimcat Networks, One Connect, OnRelay, Pactolus, PalTalk, Paradial, Peerio (Popular Telephony), Peoplecall, Personeta, Pingtel, PointOne, Psytechnics, Quintum Technlogies, ring2, Sansay, SentitO, SIPQuest, Skype, SkyWave, Strix Systems, Surf Communications, SWYX, Sylantro, Telchemy, TeleGea, Telio, Tellme Networks, ThinkEngine Networks, Trinity Convergence, Ubiquity Software, Unimax, VegaStream, Versatel Networks, VidiTel, Vocera, VoicePulse, VoIP Americas, Volo Communications, Vonage, Voxpath Networks, WebCall World, Webmessenger, Wildpackets, Zultys.

Technology Forcast Horizon 2020

Interesting piece in today Guarding on how technology will be used in 2020

As Paul Saffo, director of the Institute for the Future in California, puts it: "What defines each decade is not a technology's invention, but rather a dramatic shift in price and performance that triggers a sudden burst in diffusion from lab to marketplace." Lunar accommodation is still at the pricey end of the market - and that is unlikely to change by 2020...

So what will change? The answer lies in the way information technology will transform our day-to-day lives. "The big trends that are going to change things are the availability of cheap sensors that provide digital data, cheap computing power and ubiquitous connectivity - the ability to connect to networks," Lewis says. "Then part of what I think people will do over the next decade is start to look for things they can automate, so you won't have to do them any more." In other words, by 2020 everything large enough to carry a microchip probably will, and from there the possibilities are endless....

Cf a previous note on three interesting innovators (Millennial Net, Dust Network or Ember Corp) of the enabling technology needed this type of ubiquitous computing.

There is a loss of privacy that is going to be very difficult for people and we haven't figured out how to deal with that," Lewis says. "But if you explain what is does, how much information it provides and where it goes - and that the trade-off is that you don't have to wait as long in line at the supermarket - then people will take the trade-off. With the right rules and regulations this sort of stuff should be more acceptable....

One of my past investment at RVC Zero Knowledge System has recently created a new spin-off Synomos to provide solutions to enterprises who faced with growing pressure to protect critical data, achieve regulatory compliance and address the rising risks of liability and penalties of non-compliance, have begun to address their ability to manage, implement and audit their compliance programs.
We will see how fast the market demand and reality evolves to match the 2020 description.

CEO Bloggers' Club

Good press coverage (in French) for Guillaume du Gardier new CEO Bloggers' Club.

Guilames blogger's Club aim to gather CEOs who believe in the blogosphere and its potential and to offer a place to share with other companies leaders the experimentation they are conducting thanks to weblogs.

The club will: 1. Create an online resource available for every visitor who might want to learn/share how blogs could take part in a company's development, or want to react on an experience that is described here. Each member has to post two notes a week to share with the other members any issue directly linked to "blog - enterprise". Then the resource of the club will constantly grow with day to day "blog practice". 2. Organize bi-monthly meetings every first thursday evening of the month in Paris to meet in off line sessions. At each meeting, two members of the club will provide further development of their own experience of blogging, which could drive us to interesting debates. First meeting will be on the 14th. october 2004.

Enigmatec closes new round of financing

Enigmatec is an interesting European software infrastructure to track who has develop a solution briging the network layer and the application layer

Enigmatec Execution Management System (EMS) is the only product that allows users to codify and automate their operational procedures in a flexible and scalable way while aligning business objectives (such as Service Level Agreements), with policy-management. EMS provides the tools to integrate technical performance measures with business performance information to effectively: Ensure Quality of Service for SLA's; Reduce Business Downtime Prevention; Better Utilize Resources; Reduce Operational Risk

Last Wednesday, Enigmatec announced that it had secured $6 million in its third round of funding for sales, marketing, alliance-creating and continued product development. Returning investor Amadeus Capital Partners led the up round, with participation by returning investor Pentech Ventures and new investor Intel Capital. In March 2003, Enigmatec closed its second round with GBP2 million, led by Pentech Ventures, with participation from Amadeus. In May 2002, Amadeus invested GBP1 million in Enigmatec's first round. Founded in 2001 by Duncan Johnston-Watt, Enigmatec's Execution Management System provides a management layer between business processes that drive an enterprise and the IT resources required to execute them. The EMS is policy-based software that addresses management and utilization of existing servers and resources, execution of workload and reduction of operational risk. Enigmatec's initial customers were large investment banks, but now it has branched out to telecommunications and pharmaceuticals

Congratulation to Duncan Johnston-Watt(founder and CTO) and Indra Mohan (CEO) and the Enigmatec team for getting this done. I have been fortunate to work with Duncan at Reuters, and he combines a rare mix of technology expertise (the guy knows Java better than most of us…) and people drives (he gets things done…) that a VC loves to find in an Entrepreneur.

VC Blogging

The Mercury News had yesterday a roundup of VC bloggers. Ross Mayfield has a more complete post of this (with a nice link to Occam's Razor's... ;-), and interesting links on the following subjects:

- How to Shop for Venture Money
- Valuation
- Valuation Methods
- Beauty Contests and Venture Valuations
- The Exploding Termsheet
- Participating Preferred
- What Does Pro-rata Mean?
- Operator or Investor VC?
- When Technology Lead Doesn't Matter
- Accelerating Acceleration
- When VC Isn't Right
- Financial Fitness for Entrepreneurs
- Running an efficient board meeting
- Putting the "Power" In Powerpoint
- Torturous Powerpoint
- Investor Rationale for Newsgator
- VC Clichés
- We don't like surprises
- Being Diligent

When to stop?

Joe Kraus posting no 2 about the history of Exite and lessons in persistance (part 2)

Fact 1: There were at least three bidders for the two buttons: us, Infoseek and MCI (they had a yet-to-be-launched web search and directory product that, I think, was going to be called Genuine).
Fact 2: We had a little less than $1,000,000 in the bank
Fact 3: We were screwed..... .

is a great reading of history and courage, it’s the kind of story telling that I, that we all love to read, a history of persistence and guts, where at the end, due to hard work, the small guy wins, an it is demonstrated that when there is a will there is a way…

As an entrepreneur I to believe that persistance is key, and will make things happend, but as a VC, I ALSO want to voice the other side, the difficult question of when to stop?

As a VC this is one of the most difficult things to do, after all we are pay by our limited partners to build companies and not to shut them down…
.... As a matter of fact, no, as director of one of our investee company, we are working for our shareholders, and we are there to maximize shareholder value, which sometime will mean stopping the business, and returning the remaining cash to the investor.

I have seen this happening in several cases like:
- Wrong market timing
Great technology, and a product that intellectually should be in demand, but no supporting eco-system, and still years away from enough earl innovator demand to support the burn-rate of the company…
- Wrong management unwilling to step-down
Poorly structure term sheet where the founding members have enough power to stay, board control, but lack any of the capability needed to build the business, and the intelligent to step down…
- Wrong business model
As the product demand increased, you realise that the cost of the infrastructure needed to support the business growth faster that the revenue of the company and that you have a money-loosing machine on your hand for the foreseeable future
Etc…

In cases like theses, the right thing to do, is not to continue, and believe that your sheer will make a difference, but to close the company down, release its people, and move on.

Making this call, knowing when to stop might be the hardest decision a VC and the board of a company might have to take, and it’s important to remember that something things do not work out....

To use a movie metaphor, while I love Hearts of Darkness (brilliant moving on the making of Apocalypse Now), I also remember that Lost in La Mancha (great movie of the disaster failing of Terry Gilliam’s attempt to shoot Don Quixote) is the other side of the same question…

Unencumbered by reality

Great story by Joe Kraus, founder of Exite of his encouter with Vinod Kosla, and what makes a great VC

While we were still in the garage (literally), we met with at least 15 different venture capital firms. The meetings we're all the same. We showed them our search technology, showed them "concept-based" search, and showed them targeted advertising. To a firm, the first question they asked was a very reasonable one: 'great stuff guys, but what's your business plan? how are you going to make money?' Of course, being 22 years old and fresh out of college we replied, 'we thought you could help us out with that.' Apparently, that's the wrong answer. Who knew?...