Seems like an over-simplification. The CTO (or technical cofounder) is a key piece of the machine that enables you to iterate quickly to find product/market fit, and thus get the sales. This article almost seems to advocate not bothering with a good CTO and getting a good sales guy instead... but if you do that, you'll have a whole different set of problems.
So the boring, but correct, take-away, is: You need a great CTO - and you need a great CEO who can sell the product too.
This isn't correct at least from my personal experience.
I've talked with a few startups that have been burning through cash because they thought they could outsource their entire technical operation to India. Some of them even tried doing it a 2nd time after failing the first time. All of them were on the verge of collapse.
I think there are even more of them that never even get to the VC stage because they often times don't have any demo to show.
Now let me see if I got this right...OP is a venture capitalist who is complaining that he regularly sees pitches where the business is not focusing enough on their market and customers.
Did it ever occur to OP that this is just selection bias? How can he comment on all of the pitches he doesn't see?
Start-ups focused on revenue have many more options before they'll pitch him. Perhaps they're self-funded. Or now that they have a presence in the marketplace, they have more investor opportunities: customers, vendors, angels, strategic partners, even banks.
Customer focused start-ups are out there. OP has just positioned himself to see less of them.
How is this guy not contradicting himself? First he says:
Convince [me] that you first know who your customer is. Second, make clear that you know how to address these customers. Finally, show that these customers will spend their hard-earned cash on your product. [..I]f I don't hear the business case, it all becomes "blah, blah, blah".
Then he says:
I believe Foursquare is on to something. Sure, it's like Twitter in that it hasn't yet figured out it's business model but I can already see there's money there. Same is the case for Twitter. Give me this feeling when I am listening to your pitch. I don't need to see the break-even point. I need to see the huge exit down the road thereafter.
Yep, that jumped out at me as well, it was like the article was written by two completely different people. Plus it barely qualifies as an article, more of a quick blog post.
OP is right, and people would see that if they weren't reading it as if he were specifically talking about YC startups.
He says: Of companies designed around the sales plan, they rarely fail because of a bad CTO. That's true. It might not seem true looking at student-run startups, but that's because most of them are starting with a product they think people want and then looking for ways to sell it, rather than starting with a sales strategy and then finding a product to match it.
The former strategy is very risky, so it only makes sense to start a business this way if you're sure of a huge exit; after all, sales-centric companies still have the option of a huge exit, but they also have the option to become billion dollar businesses without an exit. What's more, being sales-centric isn't mutually exclusive with having a great product. It's true that some great products don't lend themselves to being sales centric, but the point is that if you have the option to choose either a great product that's sales centric or a great product that's not, choose the one that is.
Now the technical co-founder is absolutely essential to starting a business where the product isn't sales-centric; the point is you shouldn't start these businesses in the first place, because given a universe with infinite number of projects you should choose the one that's both a great product and also allows for immediate sales. IMHO this is the #1 thing that's currently missing from YCs advice to startups, albeit the present advice generally works because YC specializes in a very specific type of startup operating in a very specific environment and time period. It's kind of analogous to how in the 60s they designed the product around the TV commercial, in that you could imagine people giving similar advice at the time that was too narrow but often worked anyway.
edit: And I hope I'm not offending anyone by saying this, IIRC pg actually hinted that they may have to rethink their customer acquisition advice in a recent essay.
edit2: What I am trying to say is that the YC advice is incomplete because it doesn't account for the hidden disjunction, which is the same reason most people get the Wason Selection Task wrong: http://en.wikipedia.org/wiki/Wason_selection_task
So the best business is one that 1) starts with a great sales strategy 2) is something people want 3) does not need time or money to invent 4) still has a good barrier to entry. Once you fully map out the disjunction, it's easy to see why selecting startups based mostly on their ability to make something people want may have worked great five years ago, and may still work decently today, but is ultimately flawed and will probably continue to perform less and less well over time. Now it's still perfectly fine to select startups good at MSPWing if that's who YC likes working with, but making that the most important factor for predicting success is probably a mistake; ability to MSPW as a predictor of success is a shortcut, and there's nothing wrong with using a shortcut in some cases, but what that shortcut should be changes over time.
As I understand things, a technical inability to keep up with growth in the userbase is what caused Friendster to fall behind while Myspace took over its market.
Being sales-oriented works poorly[0] unless you actually have a product people want. To win, you have to make something people want, and make sure they know they want it.
[0]I suspect this may not hold up for big corporate and government sales.
"Being sales-oriented works poorly unless you actually have a product people want."
That's completely correct, however you're still missing the hidden disjunction. The point is that you should have a product people want AND be sales oriented.
Are there times when it's rational to start a business that's not sales oriented? Absolutely. But you better have a damn good reason.
For example, let's say your product is a search engine that returns results structured in the way people learn. And what's more, it's so obvious that your search engine is better that a huge percentage of people who use it instantly switch their default homepage. Now this is a good example of when you might want to start a business without a sales plan. But starting a business like this is like planning a breech birth, so you better have a good reason.
I've seen a really bad CTO hurt a company by keeping MS Outlook running smoothly, but starving engineering resources with gimpy bandwidth and horrendous latency. Then, when engineering would complain about the network, he would attack their credibility in front of the rest of the company. The resulting talent drain was epic!
And no, this wasn't a startup. It was a 30 year old company.
maybe you won't outright fail because of it, but brilliant technical leadership is half of the difference between a mediocre company and an enormously successful one.
So the boring, but correct, take-away, is: You need a great CTO - and you need a great CEO who can sell the product too.