Q&A: Foul-Mouthed Blogger Ted Dziuba Tells Why Most Startups Fail | Hacker News Reader

Q&A: Foul-Mouthed Blogger Ted Dziuba Tells Why Most Startups Fail

wired.com

28 pointsnickb41 comments

It's strange they're surprised he's only 23. It doesn't surprise me. The facile "you suck" is the hallmark of the Beavis and Butthead 15 year olds on Digg.

Have you read any of the uncov reviews? Most of the sites they are making fun of deserve the abuse, IMO.

I think the wide-eyed optimism of most Web 2.0 coverage justifies at least one more critical perspective.

You can be critical in an intelligent way, or you can simply avoid talking about things that you don't like. No need to fling mud at them.

I think uncov is critical in a reasonably intelligent way -- albeit they use humor, rather than sombre critical analysis. The things that uncov typically makes fun of startups for -- being derivative, lacking interesting technology, not providing value to users, not having a workable business model, etc. -- are substantive criticisms.

sounds like a maddox wannabe imo

Enough to get the formula. Rush Limbaugh uses the same one. But oddly enough Limbaugh is the more intellectually honest of the two, because he's not attacking things where the right strategy is to launch with a minimal version 1.

yeah paul graham, you really showed me. internet is pretty serious business!

I can't help but wonder how this story made it to the top with the new policy and all http://ycombinator.com/hackernews.html

Will I be deemed a "bad hacker" for up-voting it?

You're rubber and he's glue, with all the mighty splendor of 4chan.

It would be surprising if he was 30. Tearing down other people's ideas is easy - it's actually what some people are best at right after graduating from college.

"All these companies will keep getting bought up, but the acquirers are not going to see great returns."

This struck me as the thing we don't talk about here--not unlike certain topics you don't talk about with old German people. Is there data on how much revenue startups bring to companies that buy them?

Another way to approach this issue: if startups are capable of revenues that justify the enormous amounts that BigCos are paying for them these days, why is there such focus on having an exit? If startups are so much better at solving problems than BigCos, why can't they solve the problem of pulling in revenue better, too? This is not meant to refute the antecedent. It's an actual question that I'd be glad to hear someone answer.

Some startups do end up with healthy revenue, but you'll notice that their expected source of revenue was obvious from the start. Not so with most YC startups.

Risk mitigation. Most large tech companies don't purchase startups for the revenue they bring in. They purchase them because they are deathly afraid that some startup will become the next big thing that renders their revenue stream obsolete. It's happened often enough that big companies have reason to be concerned: think of how the computer world would look different if IBM had bought Microsoft instead of licensing their software, or if Yahoo had bought Google when it was an early-stage startup.

If you look at the > $1B acquisitions, almost all (except Skype) have been for companies that fundamentally threaten a large company's business model. If Google hadn't bought YouTube, Viacom's lawsuit against it might have set a precedent that literally puts Google out of business. If EBay hadn't bought PayPal, it would've opened the door to a payment service that wasn't controlled by Ebay yet was frequented by all of EBay's customers; it's not a huge leap for Paypal to start offering its own listings then.

Startup founders have the opposite risk profile: their wealth is all tied up in one company, so if they don't take over the big company's market, they get nothing. Most founders would rather take their $40M payout and diversify.

So basically it comes down to risk aversion. Say that a startup has a 1% chance of displacing the big company (worth $100B for the sake of argumentation), and a 99% chance of eventually failing. For a big company, 1% is far too large a chance of losing everything, particularly when they have the cash & stock to avert the possibility entirely. For a founder, $1B is a helluva lot of money, and far better than a 1% chance at being the richest man in the world yet a 99% chance of nothing.

There's also a moral hazard issue, in that the people actually responsible for footing the bill are stockholders, yet the company executives have a far stronger incentive to not look bad by botching an acquisition than to save $1B. Many money managers hate acquisitions, because it's by far the favorite way to destroy shareholder value. But few CEOs will be replaced for making too many acquisitions, while it's almost guaranteed that they will be replaced if they fail to make the acquisition that becomes a major competitor.

A pithy way to put it might be "Big companies are driven by fear, yet startups are driven by greed." That gives both of them incentives to agree on an acquisition price higher than fair market value.

Let's assume you're right. Then essentially, startups are primarily making money by exploiting the irrationality and inefficiency of modern corporate decision making? That's not very nice. I wouldn't call it stealing, but it's a far cry from PG's squeaky-clean ideal of "creating wealth".

But I don't think you're exactly right, either (though most of your points are well taken). First of all, >$B acquisitions account for a small percentage of all acquisitions, and using them as a model isn't instructive. Do you think Google was afraid that Zenter would displace it?

Second of all, when assessing the average startup's risk profile, success shouldn't be defined as displacing the BigCo. It should be defined as making some money (say, a few million). If there is a 99% chance that nobody will want to pay for it, are you really making something people want?

> Do you think Google was afraid that Zenter would displace it?

No, but I think Google was afraid that it would displace Presently, its presentation software. Since Google wants to play in the "online office software" market, they're afraid of someone else becoming a market leader in one of the sub-markets for that and fragmenting that market before they can dominate it. They also bought one of Zenter's competitors.

> If there is a 99% chance that nobody will want to pay for it, are you really making something people want?

It's not enough just to make something people want. You have to make it better than everyone else.

This is implicit in startup advice, but it applies here because we're talking about mature markets. With startups, it's assumed that you're not going to build something someone else is already doing better, because then you're not really making something people want (they want your competitors instead). And you don't generally worry about someone coming along with a better product later, because if you have a head-start and work your ass off, they'll want to buy you rather than spend the time, money, and risk of developing it themself. However, if you then spurn their acquisition offer, they have no choice but to try and beat you, and there's a good chance that they will, because they have more resources. You've still made something people want, but then you force BigCo to waste money duplicating it (and possibly duplicating it better) instead of buying you.

So basically, the reason not to try to go for revenue is that BigCo will crush you, unless they buy you. But will they every time? In Zenter's case, it was inevitable that it would come to that. But are BigCos after every market that exists? Does Google care if $1M can be made a year from something that has nothing to do with any of their products? Duplicating your technology is worth it if a lot of money is at stake, by BigCo's definition of "a lot". But if there isn't?

I don't think you will deny that, in theory, there are markets such that if you want to capture them, it's better to go for revenue than for exit. The question that interests me, what percentage of all markets do they comprise[1]? Everyone seems to completely ignore them in favor of the Zenter way, as per pg's advice. Are they insignificant, or underestimated?

[1]Of course, there will be no precise percentage figure, unless someone can come up with a list of all the things that people want. pg's next essay, maybe?

Oh, I think the majority of markets are ones where it's better to go for revenue rather than exit. By a large margin. After all, most of the country is not Silicon Valley, and most industries outside of tech (and to a lesser extent biotech) do not support acquisitions the way tech does.

Actually, I think that even if you're going for an exit, it's better to go for revenue - at least enough to support yourself profitably. You're in a much better negotiating position if you don't need to be acquired.

But I assumed that if you're here, you're interested in the normal tech-startup lifecycle. Come out with an innovative product, grow very fast, disrupt some markets, and then get bought. Because that's what PG has experience in, that's what his friends do, and that's what most of the sites featured here do. If you're in a brick & mortar business, you're much better off talking to local entrepreneurs than reading here.

Just make sure that you know which strategy you're pursuing (revenue-based growth vs. quick acquisition), and that everything you do - market, team, capital structure, business model, positioning, technology platform, engineering process - supports it. Some tech entrepreneurs think that they'll do a technology business that's based upon slow revenue-based growth - I worked for 2 years at one. However, there's a very real risk of having the market pass you by, and being marginalized in a tiny corner where you're essentially relegated to consulting. If you're going to grow slowly, you need to pick your market niches carefully, you need a robust software engineering process, and you can't afford to enter a market with lock-in.

Yeah, I see where I went wrong now. I was talking about tech startups, but it didn't occur to me that a company whose market is off BigCo's radar is by definition not a startup.

You know, I've learned something today. Thanks for this conversation.

"Under the surface, there's nothing noteworthy going on."

So what? It's about time for some of these bloggers to stop dealing with details and start focusing on issues.

The promise of a good web app is not how pretty it is on the surface or how much technology is under the hood, but the value it provides solving someone's problem.

(By the way, his filthy language scores 0 points with serious business people. If I want to listen to know-it-alls spewing forth f-words, I'd just go to any <local bar>.)

Fixed: "The promise of a speech is not how clean the language is on the surface or how much thought is under the hood, but the value it provides solving someone's problem."

Seriously, people who swear still have good ideas. Filtering them out before you can critically evaluate them is the mark of an idealist unsuited to the pragmatism of real business.

"Filtering them out before you can critically evaluate them is the mark of an idealist unsuited to the pragmatism of real business."

On what planet?

Toilet mouth is the quickest way to demonstrate that you are not to be taken seriously in business. No need to filter. Just move along and talk with a professional, not an idiot.

What about wearing white-collared shirts and neckties? Beards? Comfortable shoes?

Uncov seems like a risky idea for a group who are about to launch a startup themselves. Sure it gets them press, but they also inherit ill will and an audience looking to call them on it when their product doesn't live up to their rhetoric.

You talk as if uncov is some kind of calculated move to get them press for persai. This kind of thinking pretty much forgets the fact that people usually start sites like uncov as a hobby, or as they would put it, "for the lulz." They probably had no idea they'd be even as popular as they are now. The amount of attention it's getting now only suggests that maybe they have a pretty good point.

I definitely didn't mean to suggest that I think they did Uncov in a calculated move for press. I don't doubt that it arose organically as you suggest. I just think it goes to show that doing something like Uncov can be short sighted.

I'm not really sure that attention is a good indicator of having a good point. Bill O'Reilly gets a lot of attention. It's not necessarily because he is making good points.

Maybe not in your opinion, but certainly in somebody's.

Its not a web 2.0 startup and I think thats the point of Persai. Persai from what I hear is going to be fairly void of the things that make up the 2.0 web dance. Its be a web application that actually does something of value for the user.

Another web application I admire is Amigo created by the folks at Carsonified (was known as Carson Systems). Instead of spooging loads of ajax everywhere and calling them platforms and technologies, actually build something someone can use.

"If the business is every widget under the sun conglomerated into this giant application, there's no real technology there. There's no noteworthy computer-science problem being solved."

try to solve computer-science problems is a reason why startups fail. it doesn't matter if there's no "real" technnology - the thing could be built in lego bricks as long as it solves a USER pain it's irrelvant.

"TD: We want to build machine programs that can learn things from information that's out there on the web. In the first application we'll come out with, you tell us things that you're interested in, and we'll continuously go out and find stuff on the internet that's related to that. There's a positive feedback loop where you tell us what you like and don't, so the machine gets progressively better in learning what you like."

That's extremely ambitious. I've seen so many implementations of what he's talking about and the results have always been abysmal. Biggest problem for these systems is that you need a lot (and I mean a LOT) of training data to start making predictions/recommendations that make any sense.

Their scope also seems way too broad.

nickb -- I have no idea if they know what they're doing, but I believe they've got the kind of large datasets you're talking about. On their <a href="http://blog.persai.com/">http://blog.persai.com/</a>; blog they describe their collection of several hundred thousand RSS feed. I'm curious to see what they come up with.

artlung, you need ranked data, not just raw feeds.

And digg is some sort of super incredible technology? what about facebook in the beginning when "under the surface [there was not much] noteworthy going on." This guy is a loser, although any press is good press I guess.

/me points to the Teddy Roosevelt quote again:

http://news.ycombinator.com/item?id=66782

the funny thing here is that ted is the prototypical candidate for ycombinator:

- technical skill: math degree, radrails - belief in tech meritocracy - ambition to solve a hard problem

thats why ted's criticisms hurt so bad -- he is slamming his peers. at least give uncov some credit for both brains and balls. life without these guys would be more status quo, which is not what any of us want.

We don't consider only technical ability. Character and judgement also count.

It's like he's thinking the same thoughts I have, except he's not afraid to be mean about it. O.o

I'd defend you Ted, but I'm too busy laughing.

http://terrychay.com/blog/article/ycombinator-xobni-party.sh...

he talks too much shitting other companies when instead he should concentrate on his startup...

by the way his idea is already implemented by others: www.feeds2.com

"All these companies will keep getting bought up, but the acquirers are not going to see great returns."

gettting acquired without the potential of revenue? show us these dump acquirers, others would be interested too!

he sais he likes companies like Joost, that have real technology behind them. Yeah, it also has 100 employees.

I'm sure he is concentrating on his startup. What's wrong if he took a short break to write Uncov article?

So what if his idea has already implemented? Google's idea had been implemented by others as well. Just that Google does it better. It's possible that his startup might do it better.

I kind of agree with you that he has no solid data on these "dumb buyers" but I think the word "acquired" should be swapped with "invested". There are a lot of web 2.0 companies that don't have strong revenue but got a lot of money via investment. Let me ask you back, can you give example of web 2.0 companies with strong revenue record?

The best deal I've seen so far was Flickr's 30-40 million dollars. The rest seem to be over valued.

What's wrong with Joost and 100 employees? If Joost really needs 100 employees, be it. They have a solid product compare to most of YCombinator companies out there.

By the way, out of 50 YComb companies out there, I only use Reddit.

I like Joost. You got that one wrong.

It's natural investments may turn out wrong. Especially when expectations are larger than the potential, which is what happens..

He can do whatever he wants with his free time. I don't care. But telling everyone that is pointless if you don't solve a computer science problem, he is really illusionating himself. And raising his recomendation engine as the ultimate computer science problem is really ridicilus.

I will put it this way:

It was said during the International Symposium in Computer Architecture this year, although, i'd be mistaken who said it. I think it was put something like this to the crowd:

"We asked teenagers to tell us what is the most significant discovery. Their answer? YouTube That's what happened, we've gone so far with technology that we neglected what is important to the people."

I'm sure there are other advances in technology that is very useful to others but overlooked by American teenagers.

If what's important is to fulfill teenagers' fantasy, all of us should build more social network a'la MySpace, Xanga and create cyberworld a'la ClubPenguin, BarbieWorlds, CyWorlds instead of helping the Health industry.

I'm fond of the quote "If you're bored, it's your fault."

I think it must apply triple to someone working at Google. I'm guessing he was eased out for surfing /b/ too much.

Reply on news.ycombinator.com