Old School meets WTF Beta

In certain areas, I'm old school and I admit it.  Shocking but let's start from there. 

In the old school land, pre-production(alpha) software was risky buggy and never put on production machines with production data. Only insane geeks trusted this stuff which was designed to smoke test features, etc, etc.  The next level after that was beta.  That only happened after self-hosted, dog food eating, etc, all happened. Lots of people gave up their hard drives for the greater good.  Then, Beta. It was tough to get into Beta programs and you needed to provide feedback or buh bye. 

Today it is WTF.  Beta label means a standard "thanks for the feedback, it is in beta" comment for feedback/bugs/problems.  People are using "beta" photo editing tools on vacation photos, for example, with no thought to the 'tough cookies, it is beta' response when/if the picture(s) go bye bye along with your hard drive.  Big companies, for the most part, still try to do some semblance of bug testing or safety checking before launching 'beta' software.  Google, Microsoft, etc, all have some level of process but, of course, it is still risky.

Little companies (or start-ups) have destroyed the word beta. Beta is now an excuse to get free publicity while being able to avoid both QA and Customer Service teams.  QA becomes the unsuspecting public and customer service is an auto-responder to every email/phone call/fax that simply says; "Thanks for the feedback to our Beta program. Although we can't respond to every item, we're listening so you keep talking.."

Because we've managed to convince every hot shot start-up that the only way to success, happiness, and money is to "get it out there and let's see", we've seriously dropped the quality of products/services flowing today and we've numbed the public to what quality is. 

Generally stuff sucks and people are used to it becomes the prevailing attitude which turns into a vicious cycle of half installed crap still lingering on a machine with an already questionable starting point.

Beta software should be declared beta when it is feature complete, run through serious QA (or closed smoke testing) and generally found not to contain obvious bugs which will likely cause serious system destruction.  Bugs happen. UI problems, yup, those as well. But locking up machines, making a mess of a registry, damaging the hard drive, and causing cancer in rats, should not be happening in a wide scale public beta.

Yeah, I got sucked into putting a "beta" app on my machine and created a nightmare of a mess causing me to basically re-build the machine while on the road.

I know better and probably deserved this up the butt treatment but still, people, B-E-T-A is not a compiled fresh daily, we kill your machine so you don't have to, adventure. 

Rebuilding my machine at 3am using a hotel business center to pull drivers off the net then transfer to my machine, etc, etc, is loads of fun. Trust me, I just finished.

<sigh>  RedEye home tonight.

Comments

[Michael Buckbee]

I still think the "get it out there and let's see" attitude is correct, but that has to go hand in hand with quality.

Better to release a version of something with 5 solid features than one with 10 flaky ones.

[Chris S]

For beta stuff that you install locally, I've realized this is the real upside of owning virtual machine software. Your only risk is a particular instance. I'd been putting off playing with MS Vista, the ultimate beta release, and realized there is zero risk if I go the VM route.

BTW, I think Google is the biggest abuser of the beta tag. What do they offer that isn't beta. Gmail still beta after how many years?

BTW, sucks on the rebuild. Ouch!

[Joely]

I always clone my boot volume before doing updates; it takes about 15 minutes to do this with SuperDuper's "smart copy."

[Keith Glover]

Good point about Google, Chris! Beta is fine as a disclaimer and should never cost a dime to the users, but use the term too often or for too long and you start to discredit your dev team.

[Shai Berger]

Beta doesn't mean the same thing for web apps vs software. Since a web app runs in the "sandbox" of the browser, there is little risk to the user and it can be put into beta earlier.

If you install beta software, your potential downside is complete loss of your computer - as Rick just experienced. If you use a beta web app, your exposure is only what info you give the app, i.e. don't trust it with the only copies of your holiday photos.

On the web, we are all beta testers all the time. The feedback loop to developers is shorter and broader (less time and more people). Net result - better apps for us all.

Oh, that reminds me... sign up for the Fonolo beta (www.fonolo.com). Never listen to a phone menu again.

[Alex Grenier]

Meh. I just hate betas. Especially the web betas. It just reminds me of those Geocities pages that always had the "always under construction" animated gif on them. Beta has become a scapegoat for poor quality, broken agile methodology, or simply the fact the startup is not earning money and so they keep the option open to remove "beta" as they ask everyone to pay a subscription, or at least keep that door open for that big day when Google buys them for a few billions. Well who are they kidding - if it's live, it's not a beta!

[Ryan Brooks]

A little late, but I'd like to comment on this:

Some of the biggest culprits of the beta crime are, as mentioned, Google and believe it or not, Yahoo. I mean, look at Flickr, a product that was in beta for a long time.

Now, I admit there are applications and products out there that use the term beta like a buzzword. As Alex said, Beta has become a scapegoat.

But I ask you and all your readers to remember that Beta is also a powerful thing. Look at Tungle, a product that launched into beta today. Startups are growing more agile and also more subject to change. Stealth mode, private beta, public beta, I can't believe that these aren't worthwhile states. If a startup is doing a beta, in my humble opinion they are deriving something from it - in most cases, an opportunity to test logic, or target a certain audience for feedback.

That said however, Perpetual Beta - something that is plaguing the web industry especially - is a serious problem. My startup is still in development, and I have a very solid plan. Private beta, targeting industry professionals I've been speaking to. Public beta, to garner feedback and do 'real world' testing. A month later, bug fixes are implemented, requested features added, out of beta and into production.

In my mind, that is how Beta should be used.

-Ryan

[Shai Berger]

@Ryan: Well said!

Putting terminology aside, there is a lot of value to the intermediate step between "this is still under the development" and "we're ready for the general public".

By putting your service online and inviting 100 investors, bloggers and industry insiders to play with it, you get a ton of important feedback and you send a message to those 3 (very important) groups that, yes this product is real, and it's on its way to a grand opening.

I'm on my second start-up and I can tell you without reservation that you would be foolish to go directly from internal testing to public launch. Is anyone seriously suggesting that?