A recurring meme here (and almost everywhere else) is the equivalent of "Just Do It". Excellent advice, almost all the time. Almost. Except when it's terrible advice.
Yes, as someone who has suffered after launching too soon, I will go against prevailing wisdom and suggest the unthinkable, "Maybe you're really not ready and can do more harm than good by launching prematurely." Just a few of the bad things that can happen:
1. People will visit once, see that it's crap and never come back again, no matter what you do.
2. You will be overwhelmed by support requirements to the extent that all development stops.
3. You will be overwhelmed by support requirements to the extent that much support never gets addressed.
4. Your calendar becomes science fiction; everything has changed and it's a whole new ballgame.
5. The stress level will become so overwhelming for some of your people that you will simply lose them. Forever.
6. If you have taken people's money and not delivered, the guilt can become so overwhelming that it cripples you.
7. Your marathon has turned into a sprint you cannot finish. You have launched and lost.
I love the idea of pushing the envelope and launching sooner rather than later. You must have real world feedback and launching is best way to get it. But launching too early early is just as bad as launching too late. So how do you know when the time is right? I don't have a definitive answer, but I do know that "your gut" is a critical input. Sooner or later, you just have to go with it.
1. People will visit once, see that it's crap and never come back again, no matter what you do.
Most of your cases are valid and important, but this point is unworthy of your concern. It's unbelievably hard to produce Internet content that is hated AND widespread. More often, crap is filtered and forgotten. Scroll to the bottom of any large HN thread to see what I mean - there's always a long tail of posts that are practically unread, some written by well-respected members of the community.
Irritating some fickle people is a small price to pay for getting feedback from someone who likes your idea.
Cuil didn't have the problem of people visiting once and never coming back again despite its amazing strides. It had the problem of being crap from start to finish. If Cuil had improved, I guarantee you most of the early visitors would have given it a second look.
"People will visit once" is a nice problem to have. Realistically, by the time you have nontrivial traction that way, your software will probably be less crappy. Plus, the kind of people who will subject themselves to sites so new they aren't on The Googles have, ahem, unique expectations about product quality.
yes. I have a site that gets enough SEO juice to have 500-1000 uniques a day for the last 8 months...and none of it from the big G. I've done zero marketing...the eyeballs just keep on coming. I think its fine if some of the first eyeballs never come back...but if you stick around long enough and keep iterating, I believe many probably will.
I've heard this called the "Netscape Effect" after NS4, which was so bug-riddled and bloated that people left for Internet Explorer in droves... it makes sense.
But some of those problems sound like they could be ameliorated by minimising features and making sure that the few features that you do implement are tested as thoroughly as possible. If someone looks at your software and it doesn't do what they want, they might come back, especially if they hear that you've added stuff. If they look at your software and it tries to do what they want then crashes, it's less likely. Similarly, if you lack features then more of the support requests you have to deal with will be "I want this!" rather than "I tried to do this and it didn't work" -- so you can send back a quick "thanks for the feedback" form email and then direct development to add the features people are asking for. And, of course, if your product does what it says on the tin (and just doesn't say it does very much) then #6 won't apply.
So agreed that "your gut" is a great input on the go-live date, but another good one is:
- Have I used it enough to be sure that I know about most of the bugs?
- Have I fixed all of the ones that matter (for a fairly pernickety value of "matter")?
(Of course, the line between a "bug" and a "feature that I really want" can be a fine one, especially in some industries and for certain users. But I don't think that's avoidable -- when you launch you are guaranteed to not have every feature that every user wants.)
I've started trying to narrow the scope of features required for launch, where the minimum set of features is not with respect to the full set I want, nor with what customers might need, nor what will differentiate me from competitors - but the minimum set that would make it have some use, to some customers.
It doesn't matter if other products are more useful; nor if that minimum set expresses the really cool goal I have, or the essence of my approach. Just that it be some use to some one.
I think of this a little bit like a strategy for proving theorems: if you haven't got time to prove the thing you are aiming at, but you need to publish something, you can always restrict your assumptions and goal to what you can manage, and prove that. Now you have a base you can build on.
Yeah, the "right time" is what's important imo. Generally speaking, "just getting something out there" is a good plan - but it's not a hard and fast rule (despite what 37signals et all might have you believe :)
You should launch early because contact with the customer is what you really need. But whatever you launch needs to work. Launch with a lot less and make sure what you launch with actually works.
Launch with the absolute minimum amount that actually provides value to the customer. This is usually a lot less than you think it is.
A recurring meme here (and almost everywhere else) is the equivalent of "Just Do It". Excellent advice, almost all the time. Almost. Except when it's terrible advice.
Yes, as someone who has suffered after launching too soon, I will go against prevailing wisdom and suggest the unthinkable, "Maybe you're really not ready and can do more harm than good by launching prematurely." Just a few of the bad things that can happen:
1. People will visit once, see that it's crap and never come back again, no matter what you do.
2. You will be overwhelmed by support requirements to the extent that all development stops.
3. You will be overwhelmed by support requirements to the extent that much support never gets addressed.
4. Your calendar becomes science fiction; everything has changed and it's a whole new ballgame.
5. The stress level will become so overwhelming for some of your people that you will simply lose them. Forever.
6. If you have taken people's money and not delivered, the guilt can become so overwhelming that it cripples you.
7. Your marathon has turned into a sprint you cannot finish. You have launched and lost.
I love the idea of pushing the envelope and launching sooner rather than later. You must have real world feedback and launching is best way to get it. But launching too early early is just as bad as launching too late. So how do you know when the time is right? I don't have a definitive answer, but I do know that "your gut" is a critical input. Sooner or later, you just have to go with it.