Y-combinator, TechStars, 500startups, and Seedcamp have been generating the most publicity, but there are many hundreds of incubators - and they range from the 'shared office space, power, network and telephones' model to the full on 'platform' , where the startup teams build the product, while sharing access to everything else a company requires to operate.
This is one of those topics where every time I look at Twitter or my RSS feed another worthwhile link appears - so post, and update later.
Incubators, or accelerators, have been around for a long time. Y-combinator and TechStars, in particular, have demonstrated repeated success as measured by 'graduation' from the program - companies which get have used their initial funding to demonstrate traction, and which have been acquired for useful multiples of their initial valuation. These two, like Seedcamp in Europe, are focussed on high volume consumer services delivered using the Web or as a mobile phone app. In the networking business we call these 'eyeball companies' - part of their value is their reach to many millions of consumers, some fraction of whom pay for their use of the service, and most of whom will consume advertising along with the service - often the advertising is the primary revenue generator.
Many necessary products, including the infrastructure over which these services run, are not amenable to this model - they require more investment, building hardware, and sales through channels with some degree of customization. I've been pursuing the question of whether an incubator would be a useful model for bringing together small companies with lower volume but high value markets who could usefully collaborate, while sharing a 'platform'. One example of this platform implementation is Innovation Works, in Beijing.
The Accelerator Group describes the operation
" Innovation Works houses 400 young people working away at building a wide range of companies. About 30 of these are directly employed by Innovation Works and are called the ‘platform team’. These folks (average age I guess about 27/28) are business, finance and marketing graduates whose job it is to attend to the formation, funding, administrating, recruiting for the start-up companies and helping them with their ‘go to market’ plans. In fact, they do everything other than build the product itself."
16 June 2011 Updating (again) to add YouWeb to the incubator list.
21 June 2011 Updating to add TechCrunch article on Kaufman Fellows Research study on European incubators, and Mark Littlewood's discussion of the same topic at NESTA today.
28 June Updating to add links for the newly launched Appsterdam community for apps developers to self organize in Amsterdam.
12 July - Three additions - AngelPad (Xooglers, in San Francisco), Tandem (not the computer company) Entrepreneurs in Burlingame, CA, and PwC's accelerator for Luxemburg, using Plug and Play as the Silicon Valley base.
Last week I attended and talked at EIE'11, the investment conference run by the Informatics department of the University of Edinburgh. The focus is on creating interest in the startup companies spun out from the ideas of the academic staff and students.
Wednesday afternoon was essentially investors talking among themselves. The most useful perspective came from Deborah Magid, Director in the IBM Venture Capital group, visiting (like myself) from the Bay Area. She was very clear about IBM's approach - they need to see $10m in revenue to consider an aquisition - they get involved with much smaller companies through the partner program and competitons like Smartcamp.
Thursday was for pitches - some in person, and some by means of a series of 1 minute videos, 1 per company, presented back to back in sets of ten. There was a useful collection of data about numbers and amounts of investments into Scotland by Stuart McKnight, Managing Director at Ascendant Corporate Finance. Sandy Ferguson, Senior Partner at MBM Commercial gave perspective on the lack of interaction between the billions of pounds of capital being managed by insurance and pensions companies in Scotland and the startups seeking funding, contrasting that with the investment approach taken by, for example, Calpers, which invests less than 10% of its large portfolio in 'alternatives' which include venture capital.
Nigel Eccles talked about Fanduel, and I talked about Musemantik - both spinouts from the Informatics department, at different stages of development.
There's a list of the exhibiting and pitching companies at the EIE'11 website. Most interesting to me were Contemplate - for finding concurrency issues with Java code, with two large London investment banks as initial users; and D-light, who have developed a modulation scheme useable over LED lighting - think wifi, but using light instead of wireless frequency.
Walking by the ocean, barefoot in soft damp sand, watching curlews and plovers running into the edge of the water - that's an experience.
An evening at a friend's house, admiring the house, how the architect had designed it to use light and space to maximise the ocean view and privacy; meeting other friends and former colleagues, survivors through divorces and surgeries, catching up on each other's lives - that's an experience.
Facebook helps - with people who post regularly (or who blog,or tweet, but not many do) there's much more current shared context, a frame of reference which broadens the base for conversation. Geography matters much less, personal contact is easier, warmer, more quickly renewed.
Listening to and watching music is an experience too - one for which it is easier to have small samples,in between necessities. Sharing changes and improves the experience.
Then there are games, which simulate experiences - music, used well, can elevate the game experience.
Musemantik, which makes software for adding appropriate music to games, and in which we have an investment, is raising a seed round. If you might be interested in participating, contact us at email@example.com or firstname.lastname@example.org
Whisky in two successive posts .. Last week I went to the Tartan Day reception (part of Scotland Week), held in the British Consult's residence in San Francisco. Events there are very civilized, carefully catered, and Diageo generously provides single malt whisky. Food included haggis, served as finger food, in small pieces in the bowls of Chinese soup spoons.
There were some potentially useful conversations - talked to a couple recently moved from Edinburgh - he's managing Asian equity funds for private clients - and to someone in the high end optical component business. Also exchanged cards with people from SDI (our hosts) and UKTI. I'm looking for more ways to help with Startup Britain than answering questions on Quora.
To get there, I walked across the city from the Caltrain station at 4th and Townsend.
Despite visiting the city fairly frequently, I've never had the sense of appreciation and understanding for it that I do for London. The usual way to improve one's sense of a place is to walk it - often the City is either too cold and dark, or too hot and sunny, for that to be pleasant - this was a dry sunny late afternoon with a brisk wind. Conclusion - San Francisco has some fairly seedy areas; in the nicer residential parts there are very few cafes or good quality retail stores - it doesn't at all compare with Knightsbridge, or Stockbridge in Edinburgh. There are very few good views, despite all the hills - the streets are built up preventing long aspects.
Otherwise know as making investments in areas where you actually know something about the industry, market, and users.
Continuing the theme of summarizing from Silicon Valley - one of last week's memes was 'Angel List makes it too easy for investors to "Spray and Pray" ' . Dave McClure (Master of 500 Hats) is a founding partner at 500 Startups, an internet startup seed fund and incubator program in Mountain View. His post from yesterday collects links, and explains with typical Dave attitude why value added investment is important, and why a background in engineering, building products, marketing, getting customers and revenue is something investors should have.
".. a quantitative, high-volume investment strategy filtered based on reasonable assessment of team, product, market, customer & revenue along with domain-specific expertise, and selective follow-on investment with incremental knowledge of company metrics and progress CAN result in good outcomes.
or at least i hope so, because that's what i'm doing."
Posted here because we agree strongly with this thesis, and are making domain specific expertise available to people who understand that they need it.
There have been several new guides to the Valley posted recently - collecting the links here.
The Valley is a virtual place as well as a real place - and it believes strongly in itself. Some days it feels like a collective self-delusion - if all these people didn't believe so strongly that they could build huge user communities and that investors would value what they built, that the place would dry up and blow away.
Robert Scoble has a good summary of the places which capture the origins and recent history. Steve Blank lists more of the places where you can get a current sense of what's going on, if you have enough context to understand what you are hearing.
The Valley has about 10m people in it, but it's not completely built up. There are parks and open spaces, though they aren't as convenient as the London Royal Parks. If you are interested in living here, it's worth driving to and walking in some of the residential areas - having first checked Zillow, for example, to understand which might be in your price range. For certain areas, house prices are more affected by Google and Apple's stock price, or the secondary market in Facebook shares, than by the number of repossessions.
If you are actively looking for investors, then AngelList and CapLinked are required reading.
Last week's Nanog meeting was in Miami, with Terremark as sponsor, again. (Terremark is about to be acquired by Verizon).
Matt Petach took detailed, almost verbatim, notes for much of the meeting - like me, he was listening and watching the video stream, not attending in person.
The highlights, from my viewpoint :
Distributed Denial of Service attacks are ramping up again, according to the Arbor security report. There was at least one attack (DNS reflection) at 100Gbps during 2010.
400Gbps (16 x 25G streams) is the next target beyond 100Ge
The second Arbor talk was about mobile networks - after the RAN and the backhaul, the traffic from them runs over the same infrastructure supported by Nanog operators, but there is very little
cross over in interest. 75% of mobile operators say security is poor, bad, or non-existent.
Jim Cowie (Renesys) ran a panel sesion in the afternoon which reviewed mobile network operations, and in particular the GPRS roaming exchange, modelled after AMSIX, and the proposals for IPX (not a Novell protocol, but a private over the top network for mobile traffic, and the issues with it for voice)
Jan Schaumann, Yahoo! described their L3 load balancer (L3DSR) development - Yahoo Mail is using it, and they will roll it out to other properties. Twitter is using this too - Yahoo will opensource
the kernel modules and iptables(8) plugin available for Free BSD and RHEL. A10, Brocade, and Citrix have built hardware support.
The silliest presentation was from Neil Farquharson, Alcatel-Lucent - his first time at Nanog, used the opening of the film Bambi (he'd been watching it with his kids) as an analogy to justify a unified control plane across IP to optics. He was encouraging people to come to a BoF at Denver.
IPv6 Deployment experiences wasn't streamed, though there are slides from Level3 (deployment experience, Hurricane Electric (deployment history, details), LACNIC (v4 runout effects), and SpaceNet AG (v6 routing table).
IPv6 day discussed - Yahoo!, and Akamai. There's a lot of brokenness .. the Day is an attempt to get a better idea of how much brokenness.
Shyam Mani, Mozilla, describes their experience in turning on DNSSEC, including the initial breakages.
BGP routing: Rob Shakir on his IETF draft which aims to reduce the impact on service availablity of session reset.
avoid sending notification in first place (avoid teardown!)
recover RIB consistency (something went wrong, but can we continue functioning?)
restart BGP hitlessly--avoid forwarding impacts while getting out of error condition
monitoring--as protocol gets more complicated, need to make sure it's still doing right thing
Route Flap dampening revived - Lixia Zhang presented her students' work - suggests decoupling flap detection, flap suppression.
Jim Cowie had a nice animation of how the Internet in Egypt disconnected - it reconnected again 5 days later.
Discussion of IPv4 runout.
Nanog 51 Agenda with links to slides and video