I was really hoping I'd get asked to sign-in before I could read that article.
More to the point though, twitter's issues tend not to be the kind of thing you're looking for in beta testing. You can't test scale with a limited number of users by defintion. A service which still loses data in corner cases, exposes private information unintentionally and so on is better off in an acknowledged private beta than suffering the incredibly bad press and resentment of users which will arise in these scenarios.
Also, not all software is built equal. Twitter is strongly affected by network effects, the more people using it, the more valuable it becomes, so getting people signed up is more important than ironing out every bug. However the likes of gmail is weakly affected as you can still email all of your mates whether they use gmail or not. Spotify's value doesn't particularly increase when new users join it, but it does when it's catalogue is expanded.
Gmail also used invite codes, which - at the beginning - caused false scarcity, making getting a gmail invite code itself valuable. That's markedly different than just closing off unapproved signups.
Not all of us actually want press before we create something usable enough to get a positive review. I can understand that a writer would want an all access pass, but to us, these products are our life (or at least a big part of it). For the people who want to get press, I'll say it's really hard getting on TechCrunch unless you have invites and a strong product. Usually you need both, but when you go public expect press to be less interested in general. Far less interested. They usually want their big story, and then they're moving on to the next new thing.
People only love the fail whale because they don't see it as often these days. Sort of like how to look back on old events that may have been painful but nonetheless provide you a sense of nostalgia.
Also, I find it ironic that this article is hosted by the NYT, who wall off their own site and force users to give up personal information in order to access certain content.
So here’s a suggestion to sites still hiding behind the beta wall: Throw open the gates. Get your own Fail Whale ready. And please, let me know when I can link not to your invite-only beta program, but to your cool, slightly buggy site available to everyone - now.
This is pretty much exactly what I did when I was starting my website. Instead of locking everything down, as soon as I had a reasonable working beta, I opened it up for free to anyone that happened along and wanted to sign up. Because I didn't advertise, most of those early free signups were either referrals, or random happenstance.
Personally, I think as soon as you've got a reasonable working beta, you might as well open it up and let people try it out.
Generally I agree for small startups, but some of these well funded startups aren't merely thinking about 'surviving', they have their eyes set on 'global domination' and therefore, a mistimed launch could spell disaster from a PR standpoint or could tip off another well funded competitor before they are ready to compete.
Do those type of startups exist? Certainly there are people that think a mistimed launch or tipping off competitors would be a disaster for their site, but they're probably wrong.
Cuil had more problems than just 'timing' though. They were trying to be a google killer. That requires a fundamental paradigm shift which their product simply wasn't offering.
They could have timed their launch to the precise nanosecond. It still wouldn't have helped.
It could be interesting to have a presignup for a closed beta, but with a precise promise of how long it will be before you get in -- like the 'call hold wait estimate' some phone-in support lines have.
This answers the user concern that the wait might be nothing or forever, and gives the team another variable to play with to control the rate of signups and appearance of 'exclusivity'.
In this article, Mr. Boutin reveals his complete ignorance of systems engineering. Whether it's websites, software products, cars, blenders, or even packaged food, pretty-much any remotely novel product marketed and sold for use by everyday consumers goes through a "private beta" phase to weed out "bugs" in the system.
I suggest that Mr. Boutin learn a modicum about manufacturing and consumer product life-cycles before further attempting to practice journalism on the topic.
The really sad bit is NYT actually paid money for this "content" too.
The "private beta" is a web 2.0 trope that's going to disappear in web 3.0. It's something you can afford to do when you've got a rich sugar daddy, but when you're trying to make a profitable site on your own dime, you've got to start early.
Yes, I know about the testability problems of community systems -- I've lived them. Web 3.0 finds answers to them. Web 3.0 succeeds where Web 2.0 failed.
More to the point though, twitter's issues tend not to be the kind of thing you're looking for in beta testing. You can't test scale with a limited number of users by defintion. A service which still loses data in corner cases, exposes private information unintentionally and so on is better off in an acknowledged private beta than suffering the incredibly bad press and resentment of users which will arise in these scenarios.
Also, not all software is built equal. Twitter is strongly affected by network effects, the more people using it, the more valuable it becomes, so getting people signed up is more important than ironing out every bug. However the likes of gmail is weakly affected as you can still email all of your mates whether they use gmail or not. Spotify's value doesn't particularly increase when new users join it, but it does when it's catalogue is expanded.