(this is page 91 of 121)





Latest

Through Marc Andreessen I recently came across Stephen Wolfram's Y Combinator talk which is a great read for both fans of Mathematica and those just looking for great startup advice.

The money quote:

OK, so what about all the business school stuff? I must tell you a bizarre thing about our company: I believe it's still true that not a single person with an MBA has ever succeeded with us. Probably that's because we're really not a formula-run kind of place. We insist on understanding things from first principles. Which is good if you're trying to do things for the first time. But a waste of time if you're just doing things that have been done lots of times before.

I've always thought that running companies is pretty much common sense. It's stuff that can be figured out just by thinking, practically, about things. And knowing a certain amount about the world.

Now, there are lots of smart people who are great at their specific areas. But somehow they don't seem to engage the thinking apparatus when it comes to other things. And that's fairly crippling in trying to run a company.

I love to see a scientist talking about first principles in describing the entrepreneurial challenge because it rings so true to a lot of the hard thinking you have to do in the process of creating something new. I was talking to a friend while out in California this week and mentioned something about those times at work when you have to sit down to do the hard thinking. She looked at me as though this was somehow an optional part of most jobs which, when I thought about it, is probably true.

But thinking from the ground up is definitely a must in new venture creation. An example: a lot of the VCs that I run into can often be classified as "momentum investors" in that they like to talk about funding the YouTube of X or the Facebook of Y (usually after these companies have had successful exits or massive valuations) without really thinking about whether the translation to X or Y actually fits from what Wolfram would call a "first-principles" analysis. In contrast, the thinking VCs, like my friend Nick Beim from Matrix Partners, think through the idea and the business model from the bottom up, starting from the individual motivations of the customers/users/etc. and building up a micro economic model that fits the behavior. This is usually a much better investment of time than reading "Momentum Weekly" on the crapper or looking through sector sizing reports to find the next high growth segment that the other 1000 investors doing the same haven't already found. And it is anathema to the top-down astronauts (often self-identifying as "pattern recognizers") who must have been the kids in kindergarden jamming the Fisher Price square pegs into the round openings.

Incidentally, I met Nick while doing Tabblo which borrowed its first office from extra Matrix space. While he was not our board member directly, we spent many an afternoon chatting about the fundamentals of the consumer Internet and I often came away thinking that I needed to do more hard thinking about some of our underlying "first principles." Nick also introduced me to a great mentor and now friend, Reid Hoffman, who is his entrepreneurial doppelganger— intellectually honest to a fault, analytical as hell, and never afraid to say "it might work but the concept sucks." Always refreshing, though sometimes painful.

If you're thinking about a startup, in a startup, or even in a big company in a group looking at a new product or market, go read the Wolfram piece— it might help keep you honest next time you get sloppy in a strategy session.

Comments Permalink


I remember the first computer I was given as a child as though it were yesterday. When I was 6, my dad bought me a Timex Sinclair ZX81 which was this funky little computer that had a built-in BASIC and BASIC statement macros that you could access through a modifier key. At the time we had both an Apple ][ and an IBM PC at home but that Sinclair was special to me because it was all mine. I would plug it into an old black and white television set in my room and go nuts programming simple text input programs that made me feel as though I was denting the universe.

Today One Laptop Per Child went live with its "Give One, Get One" program— a great hack of the idea of getting an OLPC for yourself. By paying $400, you actually get two— one that gets shipped to you and the other that ends up somewhere in the third world, making a difference for someone who would otherwise never get a computer.

I'm fairly sure that this is too advanced for my 5-year old, but I just can not resist getting this as his Christmas present. For starters, it's a great cause. The machine is also likely to be a collector's item sometime down the road. And maybe, it won't be too much for him; after all, we've got the benefit of better software, richer interfaces, and a machine that sounds like it might just live up to all of the hype.

Comments Permalink

Dave Thomas, the technical world's version of Mark Twain, has a nice short article "Programming the World in a Browser Real Men Don't Do JavaScript Do They?!," which covers how Javascript is coming out on top of the dynamic language pile and will probably be one of the most popular computer languages in use going forward. It's funny to see the parallels between it and C: how they were both languages created to solve a specific problem (script the browser for JS, write UNIX in C's case) that exploded as their underlying hosts became more pervasive. The lesson seems to be: if you want to write a long-lasting meaningful computer language, get out of the ivory tower of ideal types and find a host that is just going to take off— in fact, the model could be called "Languages as remoras."

In my new job as member of the CTO's office at HP, I am supposed to have opinions on all sorts of "where is this going?" questions, at least from a technology perspective. One popular topic these days is browser-based runtimes for building the next generation of applications or: who will win the death march between Flash/Flex, Silverlight, Java, and DOM scripting with Javascript?

I should note here that I am pretty awful at doing this part of my job because I find it hard to get to an answer I believe from abstract principles— that is, if I haven't worked myself in the specific platforms, it's hard for me to get to a credible opinion on who is likely to win the hearts and minds of the next wave of developers.

But despite who wins the runtime, it is worth noting that Javascript (or its "Kentucky-cousins" ECMAScript, Actionscript, and even VBScript) all underlie most of the solutions out there as glue language at minimum, and the core language in some. So in some way, shape, or form, Javascript is here to stay.

Comments Permalink

Andrew's Monkey

Andrew got this monkey from his grandfather (who doubtlessly got it after giving Stanford money) that I've been staring at while changing diapers over the last two years. The wide-eyed stare combined with the Stanford t-shirt always reminds me of the heady times during the end of the last decade during my stay there: people in the engineering school were awash with the boundless optimism that is so endemic to Silicon Valley, and over at the business school, just about every embah was convinced their next job would be their last. When friends visited me, they would inevitably point out how foreign the culture felt, not because the crazy dot-com fever wasn't raging on the east coast (it was), but because everywhere but in the Valley, it was somehow tempered by that east coast stoicism (or is it cynicism?) that keeps people from really believing that 60lbs of dog food shipped overnight for $19.95 was really the beginning of something new.

I'm thinking about this because tomorrow morning I'm going to give a talk to the Stanford Alumni Club of New England about entrepreneurs outside the Valley, or more specifically what the key differences are between the east coast and west coast when it comes to starting companies. I'm not sure that I am particularly qualified to do this— after all, despite having worked at 4 startups, none of them were actually begun on the west coast. So the talk may instead be more of a "grass is greener" kind of thing. In fact most of what I'll be talking about is how at Tabblo we were always of the mindset that our next round of funding had to come from a west coast fund in order for us to stay relevant as we scaled. Having had six months of time post-acquisition to reflect on this, I'm not sure I completely agree with our original sentiment. But before I get to that, at the cost of spoiling tomorrow's fun talk, here is what does blow about entrepreneurship on the east coast:

1. When the 495 corridor lost the reigns of the technology-led revolution that started through the mainframe, continued through the minicomputer, and eventually transitioned to the PC and the Internet, we lost a lot more than an industry. We in fact lost a key ingredient in what the Economist has come to call an "economic cluster:" the combination of solid technical research universities, a young workforce, a risk-seeking set of capitalists, and what we lost here after Digital and Lotus navel-gazed their way into extinction: large successful industry bellwethers to train the next generation of entrepreneurial managers.

The hardest part of embarking on a consumer Internet startup here in New England is finding wealthy veins of talent to mine out of big companies that provide relevant experience sets. From my non-technical entrepreneur friends I often hear about how hard it is to find class-A engineers that know "web stuff," and we ourselves at Tabblo had a very hard time finding good direct marketing talent that understood how factors like viral adoption could be weaved into a coherent user acquisition plan. Both skills can be learned by those who are really talented, but this takes time and discipline— something is hard to cultivate because:

2. Thanks to the more conservative nature of investors here, ventures in the consumer Internet space often fall prey to the business equivalent of premature optimization, favoring getting to revenue at the expense of adequate distribution (users) or product refinement. I don't know that I would go so far as to espouse the Y-Combinator idea that you just need to "make something users want" and everything else will take care of itself— in fact if you've taken venture capital and are expected to deliver venture returns, it is irresponsible not to understand what the path to positive cashflow is, and to be testing the key assumptions at every step of the way. But an over-emphasis on this can lead to a dangerous situation where amidst slower growth than expected (which happens to just about every startup I've known at some point), the management team gets distracted by the "monetization problem" just to focus on something that might in the short-term appear to be more directly controllable. And when you've got a board of investors that encourage this trap, things can get ugly quickly.

Incidentally, the VC fund which we raised our money from at Tabblo, Matrix Partners, and our board member David Skok were A+ at helping us to avoid this trap. David was always pushing us to focus on solving the distribution problem at the cost of prematurely optimizing a business which would not at that point not have been at scale. Revenue is important, as is understanding the drivers of the business, but I've seen way too many entrepreneurs prepare for board meetings replete with spreadsheets and powerpoints that are more fitting of HP's printer business than of a rag-tag bunch trying to find a market with their product.

Both of these shortcomings can together create a vicious downward cycle that takes anyone who is not sitting on top of a golden egg idea down quickly.

Now what do we have on the flip side? The short answer: lots. Every other element of the Economist's cluster abounds here— universities, youth, capital— and we've got a whole bunch of other things to boot: access to the media companies from NY (who thanks to Google and Apple are now paranoid of anything that smells of the Pacific ocean), and an undeterred willingness to tackle really hard technical problems for periods of time that would seem like Paleolithic eras on the west coast.

But best of all, the best thing about starting a company that will eventually need regular users to scale (which is the case with all consumer Internet businesses) is that we are much less subject to the echo chamber effect of the Valley. In the Valley everyone is twittering, sharing links on Delicious, digging articles left and right, and uploading pictures to Flickr from their super phones, but the rest of the country is really not quite ready for a lot of these applications. And the sad part is that most of the companies that I've seen started appear to be aping a lot of these initial Web 2.0 experiments instead of trying to think about how to move the adoption curve back into the mainstream.

To be sure, there are some great companies that burst out of the echo chamber and into the mainstream from the Valley: Google, eBay, and YouTube strike me as three really great examples. But monocultures can be very self-reinforcing for most of us, for both good and bad. When I first came to the States, I was a fairly ok student but had the good fortune to go to a private school where there was a strong monoculture of academic achievement. And guess what? It worked its magic on me.

In the same way, I'm sure that were I on the west coast, I would probably not have embarked on Tabblo, worrying about moving bits to atoms, and building better tools out there for folks who just wanted to feel more creative. Instead, I might have been more willing to drop vowels from the name (Tbbl, Tbbblz, ...) and gone with some sort folksonomy-based social platform for content digestion. Over RSS and ATOM, of course.

I love California for its bright-eyed optimism and willingness to experiment (back to the monkey here). For sure. In fact, I'll be surprised if I don't make it out there at some point for more than my 6 day/month average. But for where I am now, I'll take a pass from the monoculture for a while and think through some of what makes the consumer Internet work for the rest of the non-early adopters.

Comments Permalink

I read this weekend that in Japan consumers are slowing their PC purchases in a major way— eschewing them for the likes of iPod Touches, smartphones, and other portable devices that can bring the Internet to their pockets. So much so in fact that major web application vendors are targeting the small screens ahead of the big PC-connected LCDs.

This could just be about the death of the current PC form factor, which in space-constrained Japan might be coming on earlier. And in fact, it would seem that the Japanese love for small computers is extending down in the laptop-for-everyone sweet spot of $300 that OLPC and others are pioneering. But it could also be that the consumer computing platform has now been defined: web applications through the browser that let you do content-consumption things, buying things, social things, and in rare cases, authoring things. With gaming as the one big caveat, it seems that gone are the days where application developers talked about using the GPU to deliver amazing interfaces, or the raw processing power of the PC to do crazy magic around speech/gesture/video recognition, etc. And so cheap access devices are ruling the day, with the feature set being around how well they can support 3G/Wifi (for connectivity) and the web browser.

This is good for us but bad for Microsoft, Dell, HP, and even Apple. When the spec for an application or platform freezes from a features perspective, it's bye-bye hypergrowth and hello slogging it on price/quality/colors deathmarch. Apple gets to avoid this a little more than the others (though not much according to the article), but for all of PC folks it will be ugly. Especially given how much of a device form-factor renaissance we seem to be seeing (iPhone, Nokia N95, Nokia 810, Googlephone, etc.).

Fast forward to Open Social, this Google-led effort to freeze the spec of social networking core capabilities that everyone is going crazy about. For starters, I agree with all of the folks who say that consumers vote with their feet irrespective of standards and that these days they are all voting for Facebook, so I don't really see what the big game-changer is in putting out a standard for widget developers. More importantly though, it doesn't seem like the right time to do it, unless of course, it is an attempt to take some wind out of Facebook's momentum.

The most interesting thing about the "social network platform" is what Facebook calls the Minifeed (Open Social calls it the "Activity Stream"), an aggregation of all of the activity that is going on between connected members of a social network that seems ripe for data mining, advertising, and best of all, experimentation around how it can be used to help people find stuff that is highly relevant (including today's killer app, helping people find each other). Open Social's Activity Stream API seems pretty simplistic and maps directly to thinking of each person's stream as an RSS feed that you might want to read or potentially jam items into. And while this may cover most of today's use cases, I'd be surprised if the tightly integrated social network sites (like Facebook) don't find a whole load of more interesting uses over the next few years. One could argue that their continued dominance sort of depends on it, especially as they build out an advertising platform capable of supporting their crazy valuation.

Don't get me wrong, standardization is a good thing, even if it implies a sort of freezing that tends to kill product categories. I just wonder whether we're quite there in the social network space.

Comments Permalink