Couple problems:

1) The burn rate of a software business with 15 employees is likely to be at $1.5 mil a year. Engineers cost more than line cooks or lawn mowers. Can you think of a software company that reached $1.5 mil in year two?

2) While running this small business, in the Valley you're competing with rocketships, which are Square and Uber at the moment. A rational engineer considering your company versus one of the rocketships is going to sense there's more upside for him at a rocketship company than a company with small business vision.

Don't get me wrong, there's a bunch of software-based small businesses (consulting shops, hosting companies, Web design firms) that are successful, but employees know that very rarely anybody outside the founding team will hit it big money-wise.

Just why is an information technology startup with explosive growth potential, early on with just a few employees, as described so vividly in Altman's lecture, usually so different, say, in level of desperation, from most of millions of US small businesses? That is, I'm not getting the solid reason for so much desperation.

On your 1) and a startup being a small business:

As a side point, for 1) your $100 K per employee per year might even be too small. I'd guess that in SV Google spends $300+ K per year per employee.

In your 1), you are assuming a lot of burn rate. Yes, I know; I know: Twitter long ran with rapid growth in usage, i.e., traction, with low or no revenue but recently, from some VC Fred Wilson comments, worked on revenue and is doing well with it. Maybe SnapChat and Instagram burned a lot of cash with low or no revenue. And maybe Square, Uber, and some others have high burn rates. And recently VCs Bill Gurley and Fred Wilson, and, now, A16Z, are complaining about high burn rates. I know. Still, from nearly all I've seen, coast to coast, "pre-revenue" usually means $0.00 equity funding.

So, connecting with my point above, maybe the offices of Square and Uber have some high level of desperation as described in Altman's lecture, but from all I can see for information technology startups with explosive growth potential and just a few employees, that such a high level of desperation, so different from nearly all of millions of US small businesses, is necessary or usual seems difficult to believe. Instead, with just a few employees, in nearly all respects the level of desperation should be just ordinary for small businesses.

Then there is the issue of burn rate and, thus, usually, VC funding:

from what I've seen, for a startup writing software and pre-revenue, about the best they can get from a VC is either "it appears that you are still in heavy development" (i.e., not a candidate for equity funding) or "when you go live with your service we'd like to 'play with it'" meaning that nothing prior to that is worth equity funding.

That is, VCs want to see at least a product ready for usage. That is, they don't want to pay for software development. That is, we get reminded of the Mother Goose story of The Little Red Hen who got no help or interest from anyone until she had hot, fragrant loaves of bread fresh from the oven and lines of eager customers out front.

So, what is a founder of an information technology startup to do about equity funding? A big point here, at times also mentioned by VCs, is that now the expenses, at least the capital expenses, for an information technology startup can be so low that really, by the time VCs are interested, the startup should be within just some weeks of getting users, ads, revenue, earnings, and enough in earnings to have a nicely successful small business with plenty of cash for organic growth and low to no need for equity funding. So, from all I can see, usually by the time VCs are willing to meet, there's a profitable small business or nearly so.

E.g., Menlo Ventures told me that they want to see at least 100,000 uniques per month. Okay, for my Web site that would be, say, for each unique visitor, an average of, say, 8 visits a month with an average of 8 Web pages seen per visit, with an average of 5 ads per page. So, if take the Mary Meeker KPCB estimate of $2 revenue for each 1000 ads displayed, then get monthly revenue of

100,000 * 8 * 8 * 5 * 2 / ( 1000 ) = 64,000

dollars for my solo startup. Moreover they want that traction growing quickly. At that time I'm looking for equity funding? I doubt it! Even if the $64,000 is too high by a factor of 10, still I'm not looking for equity funding; instead I'm profitable with plenty of cash for organic growth.

So, for equity funding, here is my

Point: For information technology startups of just a few people, nearly always, by the time the VCs are ready to write a first check, the company doesn't much need it. Then what's left for VC funding is expansion stage or, maybe, some funding in pursuit of the company being acquired. Or, again, I'm not seeing as very common Altman's description of VC funded startups with desperation.

For your

> Can you think of a software company that reached $1.5 mil in year two?

I'm not sure what Microsoft, Plenty of Fish, etc. did already in year two. Still I'm lost on your point here: You seem to have an implicit assumption that an information technology startup with explosive growth potential must go at least two years burning a lot of cash instead of making at least a little in earnings, and I'm not seeing a very good argument for that.

For 2), maybe the company is not in Silicon Valley? Even if the company has only a few employees and level of desperation like most small businesses, it might still have explosive growth potential and a corresponding "vision".

You seem to be suggesting that there is high competition for good employees, say, in developing software. I very much do not see that at all: From all I can see, the US is just awash in highly motivated, highly talented, Bachelor's degree and higher graduates in STEM fields with some to a lot of software development experience who can send 1000 resume copies and get back only silence or at best some absurd questions about how many golf balls will fit in a school bus, a question "What is your favorite programming language?" with, apparently, only one acceptable answer, C++, or a requirement that the candidate have several years of experience with just the combination of software tools and infrastructure the company is using that day. Nonsense. I've seen very highly qualified and otherwise just excellent people in computer science and practical software go for years looking for jobs and be essentially totally unemployable.

E.g., there is the parody, too close to being true, "If Carpenters Were Hired Like Programmers" at

http://www.jasonbock.net/jb/News/Item/7c334037d1a9437d9fa650...

Indeed, some famous companies have a personnel policy for technical employees of promotion into management by age 35 or fired. The US is awash in people with 5-20 years of software development experience essentially totally unemployable in software. For such a person, they would have been better off, literally, at age 20 starting a lawn mowing service so that by age 35 they would have a nice business with, say, 8 crews, be immune from being fired, able to continue the business as long as they were physically able, and also able to involve their family members in the business -- make it a family business. Software developers do not have good careers, and the US is awash in highly qualified people at relatively low wages, e.g., not enough to buy a house and support a family.

The US is just awash with good universities with good STEM and computer science programs with highly talented, well educated, highly motivated students ready, willing, able, and eager to get going in a career in information technology and with no significant offers for months or years.

For my startup, I have no question that I can hire and, with minimal training, have people writing the necessary Visual Basic .NET code using the .NET Framework, ASP.NET, SQL Server, etc. E.g., they can write what is needed for customer service. They can also learn, especially with some good consulting, how to plug together servers and get them into standard racks, connect the cables, automate the software installation and configuration, set up and operate the server and network monitoring, management, and administration, develop reports on users, usage, errors, loads, etc. For the tricky, unique, difficult to duplicate or equal crucial core code, that's mine and already in good shape.

I see what you're saying, thanks for clarifying.

I think majority of small businesses are formed to serve pent-up demand, e.g. somebody opens up a pizza restaurant, just because there's no pizza joint in that specific neighborhood, but it's high on people with disposable income (and many franchise companies will actually run market analysis for you, should you decide to go that small business route). A lawn-mowing business might also start with a few friend-and-neighbor customers, and make some back of the envelope estimates on the demand and customer base.

With technology startups not only you have to build the product outright (your Mother Goose example), but then you have to acquire the customer base, so all along you're quite unsure that is achievable. As @sama mentions, you could spend a few years of your life building something, and then discovering the market doesn't need it. The realization of that weighs on founder, as by the time you've hired a few people (@moskov's quote from lecture 1), you also feel responsible for their time and opportunity cost.

It's hard to debate the point on over-supply/under-supply of software engineers without concrete data, but I think as far as your point on them being experienced in a completely different stack, a lot of startups would tend to follow the Python Paradox - http://www.paulgraham.com/pypar.html - " if a company chooses to write its software in a comparatively esoteric language, they'll be able to hire better programmers, because they'll attract only those who cared enough to learn it. And for programmers the paradox is even more pronounced: the language to learn, if you want to get a good job, is a language that people don't learn merely to get a job".

Your point on being able to assemble a team quicker/cheaper/more efficiently on relying on .NET stack is also valid nevertheless. I was just trying to guess why a certain company would choose a specific language and a specific set of tools, and require proficiency in it.

And the person asking wanted me to have experience using Python in a "production" environment. That's like asking if I've used a Toyota to haul hot asphalt, bricks, sand, and gravel. I wouldn't want to face a factor of 10 times slower in speed in a production environment.

Next I wondered, K&R C has no threading. Yes, Posix does, but, when building on very standard, portable C, there is no standard threading infrastructure. So, I looked: Right, from what I saw, on a fast look, C Python has problems with threading. No surprise. So, on a processor that can run 8, 12, 16 threads or a server with two such processors, I want to use a language that has no threading, in production? Bummer.

Otherwise as I looked at Python a little, I saw a bottle of a blend of mostly very old wine. I saw next to nothing new.

Then I found that much of the interest in Python is some packages following some package functionality (back at least to Ada). Some of the packages are good for getting and parsing Web pages as an alternative to screen scraping -- good to know in case I need to do a lot of screen scraping.

So, maybe some of the packages are nice.

So, Python looks like a promising one-shot project or scripting language. Okay. However, I'm writing production code.

Then I saw Iron Python, on Windows, apparently with good access to the .NET Framework. Okay. But Windows already has a scripting language with good access to the Windows APIs, their Power Shell. The scripting language I've been using is Mike Cowlishaw's Rexx. I've gotten a lot of good from it, but, really, I should convert over to something closer to Windows, likely Power Shell. Here I agree with Altman -- focus and don't get distracted. Well, for now, Rexx is just fine, and converting to Power Shell would be a distraction. Same for Iron Python and C Python.

For evaluating programmers by their eagerness to learn Python, sorry PG. I've learned a lot of languages. Net, net, I see that more than the language, or whatever flavor of syntactic sugar is there, is important. On Windows now, what's really important is their work for managed code, memory management (with automatic garbage collection, apparently mostly good enough for production -- "Look, Ma, no C++ memory leaks!"), their common language run time (CLR) and it's benefit in permitting programs in different languages to work closely together, and their .NET Framework. To me, those are more important than anything about C Python, and without even mentioning a programming language.

I chose Visual Basic .NET because it looks verbose, easy to write, easy to read, easy to teach and learn, integrates nicely with ASP.NET (surprisingly nicely, even without Visual Studio), compiles fast, has good compiler error messages, and is about as powerful in access to the CLR and the .NET Framework as anything on Windows. And it is essentially compiled. And it has thread support. And so far I may yet to have found an error in the compiler -- good.

Sorry, I don't want my background in computing evaluated by my enthusiasm for Python. Sorry PG --- I really like a lot of your essays but not that one.

When I teach a class in programming, I'll start with the basics -- define places to put data, allocate (and if necessary free), expressions and assignment, If-then-else, Do-While, call-return, try-catch. Then I'll teach the Visual Basic .NET versions of these.

I'm regarding programming languages not just for curiosity but as tools for another purpose, a successful business that provides as a service that improves the lives of 1+ billion people and that they like a lot, that makes money enough to do nice things, e.g., add a wing to a local hospital, help a private school, provide stable jobs, let employees put down roots in the community and improve it, pay the employees enough to have a nice house, one parent, likely the wife, stay at home with the kids, as many kids as they want, puppy dogs and kitty cats, nice vacations, private schools for the kids if desirable, college for the kids, late model cars, nice backyard parties for friends, sports, time for the family to be together, and a nice retirement. For that goal, I see Visual Basic .NET as a fine programming language.

For the guy asking about Python, I also sent him some code for two nice, original at least for me, algorithms. And I sent them a nice paper in computer science, published in Information Sciences. He and his buddies were more concerned about Python! No, I don't want to work with them. And, no, I won't hire like that.

> I was just trying to guess why a certain company would choose a specific language and a specific set of tools, and require proficiency in it.

The HR people and the suits have contempt for technical people and, really, are afraid of them and are really unable to evaluate technical people. So, as in the parody, they have little check boxes.

In information technology where software, etc., is "eating the world", the HR people and suits are total losers and in line to gather like dead flies on a window sill, fall like stalks of wheat before a John Deere combine, etc.

Thus, the opportunity is for technical CEOs. And that's why I watched Altman's lecture.

I'm not completely wrong! Commonly many people will regard any business, whatever the revenue or acquisition value, with 15 or fewer employees as "small". They just will!

And there is the recent A16Z essay that indicates that we may soon see a startup, with a solo founder, worth $1 billion. So, such an example would agree with your point. But such an example would conflict with Altmam's lecture where he is really down on solo founders!

There is an example of a successful solo founder: As I recall, early on the Canadian romantic matchmaking site Plenty of Fish was just one person, the founder, two old Dell computers, ads just via Google, and $10 million in annual revenue.

So, let's, see: Since the business was growing, maybe we should assign a P/E of 100? Since likely ballpark $9 million was annual (pre-tax) profit, maybe for the worth of the business we should multiply by 100 and get $900 million, ballpark $1 billion, from one person! A "small" business? Maybe not!

Yes, what Altman was talking about was a venture funded business with some explosive growth potential, but, still, when the company had just 2-3 people and was on the way to 15, trying not to hire, maybe not yet profitable, their value unknown except for the terms of the VC investment, maybe they are still a "small business".

Maybe a point would be that what Altman was talking about is so desperate to achieve the explosive growth potential that the office likely necessarily has a desperate air where one bad hire could wreck the company. Still, the strong conflict with so many millions of US small businesses is tough to swallow.