Hacker News Reader

The 'Hollywood Launch' can work if you already have notoriety to leverage, and already have the experience/userbase to do iterative improvement/testing internally. Steve Jobs is master of the Hollywood Launch.

If you're a small, young team making your first product, the dream of a 'Hollywood Launch' is destructive. It causes you to delay getting bulk feedback from disinterested strangers. It slows the organic learning/testing loop. It delays launch as you try to ensure you'll make a good first impression. And as Aaron notes, there's always something that goes wrong on a contrived 'big day'.

Steve Jobs and Apple on their umpteenth launch can usually keep those glitches to a minimum. (But not always: see MobileMe.) 37signals on their Nth launch of another webapp in the same mold as their other N-1 apps can pull this off, too.

But a young team launching a brand-new product will gain more from a measured ramp-up than an engineered hype-storm. They should listen to Aaron.

When you're finished with the ad hominem part of your attack, I'd just like to weigh in and say that Aaron got this exactly right.

Aaron advocates avoiding the "Hollywood Launch", marketing your product only after you've verified that "random people" are satisfied with the offering.

I'm asking you: are you saying that after Fog Creek spends 9 months building and testing a product, dogfooding it with internal users, and running closed beta testing, that you would avoid publicity for it until random paying customers started liking it?

Because I feel like I remember you making an actual movie about the last product you launched.

again, with the ad hominems!

Looked more like a question to me.

Best comeback ever (tptacek, I mean :))

I think you definitely need some balance between the two. Do a glitzy hollywood launch only after you've minimized the things that could go wrong.

This is the definition of ad hominem. For those like me who didn't know.

An ad hominem argument, also known as argumentum ad hominem (Latin: "argument to the man", "argument against the man") consists of replying to an argument or factual claim by attacking or appealing to a characteristic or belief of the person making the argument or claim, rather than by addressing the substance of the argument or producing evidence against the claim.

More on wikipedia http://en.wikipedia.org/wiki/Ad_hominem

It could be that Aaron is talking from bitter experience here.

Infogami was very much a Hollywood launch. He did almost exactly what 37signals suggested. He posted a teaser as soon as he got into the YCombinator program. He had mavens and insiders try out the product (I think; I can't imagine PG not trying out a YC startup's product before launch and showing it to his friends). And then he launched with a big splash.

And people hated it.

I suspect that product quality has far more bearing on startup success than how you launch. And as you say, the difference between 37signals and Aaron Schwartz is ability to execute. But Aaron's advice is a good way to find out if you haven't yet executed well and improve the product before everyone else finds out.

I think people hated it because there was nothing there.

He promised that Infogami would revolutionize making webpages the way that the Macintosh revolutionized the computer industry.

What he delivered was another wiki. I can't believe PG thought this was worth investing in.

Geez, that's a lot of ad hominem. Did I accidentally murder your dog or something?

I wasn't saying anything about what 37signals does, I was criticizing what they recommended. You know, in their book. That I link to in my first paragraph.

So, while 37signals doesn't break it out as clearly as you did, they do in fact recommend the same closed testing that you're recommending. For instance, in their book "Getting Real", they talk about how they do closed beta testing of individual features inside of their applications to selected groups of users, just like Google Mail does. They also publicly did closed beta testing of Highrise, their most recent application. I'd be surprised to hear that they didn't quietly test Basecamp before it shipped.

On the other point, I think I felt like in a discussion of whether we should be doing things this-way-or-that just because somebody said so, your resume was fair game. I think differently of that now. I'm sorry I offended you.

I believe that when 37signals created Basecamp it was for internal use. They built it for themselves. As time went on other small consulting firms (possibly that 37s knew) saw it and thought it was useful. From that point, Basecamp just continued to grow in popularity.

When nobody knows you or your startup do you really need a closed beta??

No, he is testing Jason Calacanis's approach to PR. By attacking Aaron Swartz ad hominen, he raises his fame/notoriety in hacker community.

Oh. I just forget Linus Torvalds called security people m________ting monkeys. It seems very hard to discuss issues rationally lately without calling names that based on some a priori assumptions of the opposite side.

He called OpenBSD people masturbating monkeys. He called security people whores. He actually had a sympathetic audience among many security pros for that statement. The whole thing is indeed a bit shady.

That doesn't make me any happier with my original comment, though.

Does this mean that the cause of hate comes from security people charges high price by hour to give people false sense of satisfactions? It seems that all people charges high price service by hours are hated in general such as attorneys and so on.

How life is so unfair and I feel your pain. But Jesus Christ loves everyone who charged clients by hour and then turns over a new leaf to wash his corpse in funeral.

Dude, I think you just wrote a Tool song!

Geez, that's a lot of ad hominem. Did I accidentally murder your dog or something?

Nah, that's pretty much tptacek's commenting style, I wouldn't take it personally.

While I agree 37signals has a much more impressive product track record than Aaron does and know quite a bit about marketing their products, the source of a statement has nothing to do with its validity. Either Aaron's argument is right, or it's wrong, independent of Aaron's status as a developer or entrepreneur.

That being said, I think Aaron is right about the proper method to launch a website, and also that if you look how 37signals actually acts, they launch products exactly that way.

I think there's a difference between the 37signals plan and the Aaron Swartz plan, which is that the Aaron Swartz plan says "don't market until you have happy customers", and then uses the spectre of product quality failures to back the argument up. Happy customers and crashing servers are orthogonal issues.

I actually agree with Aaron, but I do think his experience is relevant. Yes, he's either right or wrong regardless of his track record, but the intended reader doesn't know whether he's right or wrong, and won't know it without trying it. If you don't have reasonable way to verify whether what somebody says is true, looking at their credentials is probably the best you can do.

In practice, we do that all the time. For example, you probably have an opinion on global warming, or whether smoking is bad for you, but did you read the relevant papers, or just take the word of the researchers you thought were most credible?

In this instance, we all happen to be people who spend a lot of time thinking about startups and websites. So it's like someone posting something on global warming to a climate modeling forum - I would expect people to judge on merits not authority.

I think that's a harsh analysis and I don't think it's necessary to get personal about it. I'd say the launch of gmail was expertly done, and one that a lot of people copy.

Look at what happens when companies decide to "release" to everyone at the same time with a big PR frenzy - cuil. The companies that slowly but constantly grow seem to do better than the ones that go for a big style "launch" IMHO.

Releasing with a big style "Version 1" approach just doesn't fit with the dynamism of the web. Throw something up, get a few users, update it every day, and grow slowly but surely.

You're right: if you do a glamour launch of your product without making the effort on quality, you end up like Cuil (and I concede that Cuil is a good counterexample).

I object to the "either-or" argument the original post makes, and I object to the snide "we're only doing it this way because 37signals says to", as if 37signals doesn't have a store of credibility here that Aaron lacks.

Apart from 37signals though (Which isn't a good example IMHO - they have a rails 'fanbase'), what successful "Big release hollywood film" examples are there?

That's a good question. I'm having a slow day today; let me go look. If you don't see a response, I'll have conceded that there is no other good example of a successful "Hollywood" launch by a small startup. That would definitely hurt my argument.

[edit: my slow day got fast, but I also did have trouble coming up with a clear-cut example of a small company glamour launch success.]

Paypal?