Hacker News Reader

Comment by mattmaroon | Hacker News Reader

People make the mistake of looking at "internet companies" as a type of business (like clothing store in a mall) rather than any number of types of businesses using a common delivery service. It's no more useful to lump together every company that uses the internet to deliver/sell a service or product to customers than it is to do the same with every company that mails things via UPS.

If you define startup as anyone making money in a scalable fashion over the net, there are tens of thousands of them scattered all over the globe, including dudes who sell stuff on ebay, dudes who buy Adsense links to market affiliate products, etc.

View on HN

Like any non-mathematical word, "startup" isn't precisely defined, but common usage requires the company to be more scalable than most dudes selling stuff on eBay.

If you want to see what the degree of clustering is, the way to do it is to set some threshold and draw dots on a map for everyone over it. Tech IPOs are one such threshold; founders in the Forbes 400 is another; if you draw these maps and you find the dots are evenly scattered, you've got a counterexample.

As an exercise in research outsourcing, I found a list of 2008 technology acquisitions and paid some random person on Mahalo Answers $30 to research where those companies were when they got bought. I'm going to drop them on a Google Map, but I was quite surprised at how global they really were.

I'm also going to do it for 2007 (what's another $30?).

Here's the 2008 list. Once I do it for 2007, I'll map it out. Out of 106, 81 of them were outside of the Valley. There might be a few more, as I don't recognize all of the towns clustered around the valley. It would be interesting to plot exit events over time-- I'd guess that we're trending towards LESS clustering rather than more.

Moniker - Pompano Beach, Florida, United States Bodybuilding.com - MERIDIAN, Idaho, United States CleverSet - Seattle, Washington, United States Anywhere.FM - San Francisco, CA, United States Audible - 160 employees split between its headquarters in Newark, NJ and an office in London, UK. Maven Networks - Boston, MA, United States FoxyTunes - Israel Vehix- Salt Lake City, UT, United States HotOrNot - San Francisco, CA, United States Compete - London, UK BlogDigger - Washington, D.C., United States Auctomatic - San Francisco, CA, United States BeInSync - Tel Aviv, Israel Prospero - Burlington, MA, United States Social Platform - Los Angeles, CA, United States Pluck - Austin, TX, United States Bebo - San Francisco, CA, United States Sway - Middleton, WI, United States buy.at - London, UK YaData - Tel Aviv, Israel Weblistic - Fremont, CA, United States XIV - Tel Aviv, Israel Apertio - Bristol, UK Onaro - Boston, MA, United States MySQL - Uppsala, Sweden Trolltech - Oslo, Norway Fraud Sciences - Tel Aviv, Israel E-Dialog - Lexington, MA, United States MessageOne - Austin, Texas, United States G-Technology - Santa Ana, CA, United States Danger - Palo Alto, CA, United States Caligari - Mountain View, CA, United States Twhirl - Germany Pageflakes - San Francisco, CA, United States Sphere - San Francisco, CA, United States Farecast - Seattle, WA, United States Activeweave - San Francisco, CA, United States Fleaflicker - Tenafly, NJ, United States Expensr - San Francisco, CA, United States gBox - Cupertino, CA, United States MeeVee - Burlingame, CA, United States Inquisitor - Sunnyvale, CA, United States Cnet - San Francisco, CA, United States Ars Technica - Chicago, Illinois, United States StarBrand Media - Los Angeles, California, United States M:Metrics - Seattle, WA, United States Celebrity Baby Blog - New York, NY Snapvine - Seattle, WA, United States SwapDrive - Washington, D.C., United States Hostopia.com - Mississauga, ON, Canada Imity - Copenhagen, Denmark Rupture - San Francisco, CA, United States ZYB - Copenhagen, Denmark StarNet Interactive - Tel Aviv, Israel Plazes - Berlin, Germany Adify - San Bruno, CA, United States Personifi - Fort Worth, TX, United States Navic Networks - Waltham, MA, United States Gracenote- Emeryville, CA, United States Simple Star - San Francisco, CA, United States Diligent - Framingham, MA, United States EDS - Plano, TX Hands-On Mobile Korea - San Francisco, CA, United States B-hive Networks - Herzliya, Israel OpenAir - Boston, MA, United States Let It Wave - Paris, France Practique Associates - Bracknell, UK MusicGremlin - New York, NY, United States Skywire - Las Vegas, NV, United States Symbian - N./A Alltel - Little Rock, Ark, United States Helio - Los Angeles, CA, United States Iomega - San Diego, CA, United States Powerset - San Francisco, CA, United States Weather Channel - Atlanta, Georgia, United States HaloScan - N/A Imagekind - Seattle, WA, United States Truemors - Palo Alto, CA, United States ContentNextMedia - Santa Monica, CA, United States PodTech - Palo Alto, CA, United States jkOnTheRun - Houston, TX, United States Omnisio - Atherton, CA, United States AbeBooks - Victoria, Canada DailyCandy - New York, New York Blogcritics - San Francisco, CA, United States Ciao (Greenfield Online) - Connecticutm, MA, United States social.im - San Francisco, CA, United States Peerflix - Palo Alto, CA, United States Napster - Los Angeles, CA, United States YoVille - N/A PartnerUp - Shoreview, MN, United States Socialthing! - Boulder, CA, United States ZAO Begun - Moscow, Russia Pure Networks - Seattle, WA, United States Jabber - Denver, CO, United States Ribbit - Mountain View, CA, United States PostPath - Mountain View, CA, United States Lefthand Networks - Boulder, CO, United States Bitwine - Tenafly, New Jersey, United States Bill Me Later - Omaha, Nebraska, United States DBA - Denmark Polldaddy - Sligo, Ireland JungleDisk - Atlanta, GA, United States Wayport - Austin, Texas RuTube - Moscow, Russia Clickpass - San Francisco, CA, United States Revolution Health - Brooklyn, NY, United States socialmedian - New York, NY, United States AdEngage - Los Angeles, CA, United States acerno - New York and San Francisco based. Lookery - San Francisco, CA, United States Centennial - Wall, NJ, United States

This is not very high quality data. Extrapolating from the cases I know about, I wouldn't be surprised if half the deals on it were merely nominal acquisitions. You would have to have some threshold on the size of the acquisition for the list to be meaningful.

Looking at the locations of companies acquired in small deals is particularly misleading as a measure of the clustering of startups, because smaller deals are generally done as an alternative to series A rounds. So they're disproportionately likely to happen to startups that don't have access to funding.

For example, YC funded one startup that had an acquirer appear on the radar screen about a week after they closed their series A. That had priced them out of the acquirer's reach (for now). If they'd been located outside the Valley, the startup would not have been able to raise money so quickly, so the acquisition might have happened.

I think you've reinforced pg's point - 24/106 (~20%) is a huge effect as compared, like you've done, to every other place on the planet. Basically, from this data, if you're a tech company and want to eventually get acquired would you rather choose a 1 in 5 chance or, based on a quick guess at second-best - NY area (7 in 106) or MA area (8 in 106), a 1 in ~15 chance?

Great idea. I'd love to see 2007 as well.

"N% of startups acquired were located in city X" does not mean "startups located in city X have an N% chance of being acquired".

Take the data at face value (irrespective of whether it's the right data). More deals are going to SV and by a significant margin. You can parse causality any way you'd like (funding, connections, etc.) but that's a correlational point in the Valley's favor. Given everything else we know about the region, we don't have to work hard to see the effect.

More problematic, the data doesn't show N% because we don't know the denominator. It could be that SV is terribly inefficient at exits relative to a place like Austin or Israel. Still, the magnitudes are so different it doesn't really matter one way or another.

Yaw, I think it's a large number. 23% for one little collection of cities.

Right now, the data is really only merely interesting. The question I want to know the answer to is more related to ratios.

i.e. for every 100 startups that spins up in the valley, what percentage ever have an exit vs. the rest of the world?

It would also be interesting (as PG suggests) to screen by deal size.

Once I get 2007 and spreadsheet it all out, I might try a fiddle with a few reports.

I'd love to see an answer but intuitively it feels like you're entering dissertation territory! I wonder if the Valley is less efficient than other locales, but like Hollywood, I wonder if investors would really care one way or another.

How can you not upvote someone who spent $30 on data sorting?

<cackle> It was actually data research. The list was out there but it was not obvious where the companies WERE and I quickly got bored of reading thru press releases. I couldn't see an non-time-consuming way to get locations.

Here's the question on Mahalo Answers: http://is.gd/sCpK

I should note that the original list is NOT scientific. Near as I can tell, they pretty much made a table of the acquisitions mentioned on TechCrunch and VentureBeat, which probably do a REALLY good job of catching all of the Valley acquisitions, but probably miss a few elsewhere.

Deleted comment

JINX!

I only regret not being able to upvote this more than once.

Those metrics do not represent startups in general, only extremely successful ones. That's like saying that soccer is a local phenomenon to Europe because that's where the most successful players are (at least financially).

I'm not sure that's a non-controversial point. Dell or BetterWorldBooks & many others were essentially selling stuff on Ebay to start.

This post brings up simple & important points: (a)Of all the businesses being started in the world, only some are of the Silicon Valley cluster type. (b) Not only the Silicon Valley cluster type of business benefit from the changes you describe (again in previous essays): reduced cost of starting up, ability to grow without that much capital, & suitability or small teams.

As you mention, the definition of a startup is problematic. But maybe your definition is biased in a way that equates 'should move to Silicon Valley' with 'Startup' more then is necessary.

I meet with quite a few small businesses starting now. They are generally not the type that will be acquired making multiple founders immediately rich. But many of them are viable, wealth producing & potentially successful. Probably none of them will move to the Valley. Few or none are part of this 'startup' sub-category of new businesses.

But I think they might still be part of this same high resolution revolution. They are following business models that did not exist 10 years ago. They are (i think) proliferating for the same reasons that 'true startups' are proliferating. The are enabled by cheaper starting costs, ease of reaching markets, easy outsourcing. These are all derivative of the same cheaper hardware, free software, smaller teams, communication technology, etc.

Paul, out of interest, at what phase do you think the term "startup" changes to "business"? It's pretty clear when there's an IPO or a big aquisition, but what about companies that just steadily build their revenue like 37signals? I wouldn't call them a startup now, and they dismiss the term as arbitrary and favour simply "business".

Reply on news.ycombinator.com