How not to launch your company
blog.mutlicorp.com
blog.mutlicorp.com
Would you have signed thousands of new paying users? Ignited a media firestorm and suddenly found yourselves on the cover of Inc magazine? Would you be rolling in millions today?
Or would you have had 10,000 visits to your homepage, a small handful of signups, a few dozen comments on the HN thread (half snarkilly dismissing it, the other half arguing about some completely separate issue that the first person to comment happened to mention in passing), and a traffic spike for a day or so followed by a return to your normal levels.
Having occupied that top slot a few times, I'm sorry to inform you that it looks a lot more like the latter case. You still need to build and grow your business from square one, and a year later you'd have been exactly where you are today. A few Techcrunchings and HackerNewsings really aren't going to make a difference in the outcome of your thing.
The real difference is made by the things you did on the other 364 days between then and now. Here's hoping you executed well.
I dont get this BS about depending on a singular date as a defining 'launch'. Its a good internal goal to have, but not the ideal way you want to reach your market.
On the App Store, if you do zero marketing, you will likely have your most total downloads the day after your app is approved. The first day, you likely only get a half day, and you have to wait for full propagation on iTunes. Also, the rankings are a moving, weighted average, so unless you have a terrific spike, you need to wait for your average to start kicking in before you move up the charts.
Day 2 or 3, you will peak, then rapidly decline. So, it makes a lot of sense to time your marketing boom for the day after your approval, or rather the first full day of being live on the store. Marketing tends to be multiplicative.
People make the mistake of thinking app rankings get you sales over the long haul - it's more the other way around. But at the beginning, you get a bump, and you should use it. This bump derives from several sources:
* new list on App Store
* lots of robotic websites and Twitter accounts spamming about new apps
* the press cares about "news"
* energy from your team on a specific launch - people tell their friends and social outlets, and actual planned outreach by the company
* Apple featuring you in category or on home page of iTunes
Over the course of the first year, I saw a gradual increase in sales, presumably due to the page rank of my website increasing, or maybe because adoption of the Mac App Store increased. Now there are slight fluctuations in sales numbers, but they are more or less constant.
So I didn't get that bump, but I was lucky I made an app that was in demand.
If you did that, you'd never launch.
Market research means knowing your customers, as inside and out as you possibly can. And, if HN users represent a significant chunk of who you expect to be your early adopters, then you should make it your business to keep abreast of everything happening in the HN world. The world of stories of potential interest to HN is direct competition to your launch announcement. If you want to be newsworthy, know what news you're going up against -- especially as relates to your specific audience.
(In fairness, I don't think anyone in their shoes could have predicted Steve Jobs's death, down to the day and time. But even still, there could have been a hundred other big-attention topics that would have buried them, had they not been paying attention. Such as another major product launch, or what have you -- things that would have been slightly more predictable.)
There wasn’t anything we could do about the poor reception on HN,
but we’d been working on this for three months
expecting it to be a smash hit as soon as we launched.
That way you can make a plan that will actually fit reality and create your own growth.
Launching quickly # Launching prematurely.
I think that it takes this kind of suffering to fully appreciate the fine line between launching too early vs. too late. Sure, we're all in a rush with too much to do, but everyone should do themselves a huge favor and slow down long enough to come up with a good test plan, before, during, and after development.
I went through the same thing once and will never let it happen again. It was so bad, I actually considered becoming a cook instead. In the long run, it's just not worth it, to your customers, and mostly, to yourselves.
But having a thorough test plan better be a part of your project roll-out. No one wants to go through the rabbit holes of corporatized QA team B.S. but product refinement is a different topic entirely than the point of his comment, which was testing that product before launching it to the masses.
If you think your product won't be seen by many customers, well OK. Set it free. If you have a major push behind it, however, and know 100's of thousands of people are going to be mashing it to death in the first few hours or days, adding a time buffer for testing beyond your close family and friends is wise advice. It will save you money in the long run, too.
The alternative is to become buried under an avalanche of crash reports, bad reviews and ratings. And, conversations like, "If only we'd had more time to test this thing."
Looking at the website I am impressed you at least have an SSL certificate, cause it doesn't look like much more effort went into it. I'm even more confused looking at the about page, it lists David as being a rockstar programmer who enjoys working on difficult scaling problems, yet he didn't plan for more than a couple of users?
If you want to make your service succeed put an embedded Minecraft applet on there and have it link to some of your public servers, maybe even let the user select from a few of them. Just do the same things as http://minecraft.net/demo. Add some color to your site, your target audience isn't a bootstrapped web developer, it's a highschool kid who wants a Minecraft server. Add a big old buy now button on that homepage. And why does the Sign Up page not have the pricing options?
Is this a joke? Steve Jobs was, more or less, in ill health for years and there are enough famous people to make this advice "don't launch ever."
Somehow I don't think I've ever heard of Minefold. Sounds like something I would have been interested when I was still playing. Two years ago, I dumped a few hundred hours into Minecraft though and got pretty burnt out.
Also, being no stranger to... "problematic" launches, I can totally sympathize here.
Can't wait to see what you guys have in store with the new update, and as long as Cook or Schiller doesn't suddenly kick the bucket, I think you'll be fine with this one :-)
Sometimes you can recover without that (great job on the recovery!), sometimes you can't. If you choose your "suit" right, your chances of recovery will be better.
As I am thinking to launch an app sometime in the future, how do you test for this? This is something that has been bogging me a lot. I can't find easy to understand info on how you do this, especially running on a virtual server (so how much ram? how much bandwidth? etc.).
Try to come up with some automated test or script that exercises the full set of features you'd expect an actual user to be interacting with over the course of a session.
Then, build a harness that slightly randomises those activities, and fires off a large number of 'virtual users' against your service. Running from geographically diverse hosts will ensure you're closest to what an actual user would experience.
Actually determining what mix of operations constitutes an 'average session' is relatively difficult, but there are tools that can record sessions from actual users (look at the HAR dump in chrome, or Selenium[1] testing framework)
All of your test clients should be recording response latencies + errors, which you can later analyse to see what slows down at scale, or which critical features with unacceptable performance you need to target first.
Another thing to consider, especially during release, is that lots of users signing up and playing around is likely to be a completely different workload to regular users using it for day-to-day activities.
You need to make sure things like your transactional email notifications ('your account has been created, verify your email address, ...') can handle peak demand as well.
Lot of first time entrepreneurs learn by launching a product and getting feedback. I think there is no better way to learn than by getting the feedback from an unknown stranger.
http://www.meetingburner.com/blog/2012/10/05/a-story-of-two-...
Eg - talk about what your service actually is in the post, link to it, and lay out a call to action so you turn some of the HN traffic into actual users vs browsers . . .