Ask: How do you actually launch a beta? | Hacker News Reader

Ask: How do you actually launch a beta?

35 pointsbenologist17 comments

So recently we finished up our platform (finished in the sense that we locked the features down and cleaned up all the loose ends, as opposed to 'truly' finished).

We have a business plan, fantasies about investors, dreams of adding and doing 1000x more than we've already done and conquering the world, bla bla bla.

So how do you "launch" a beta, what do you to do interest the relevant startup blogs and generate some publicity and attention for your freshly baked business? Or do you just skip that and quietly launch while you go fund-hunting? What do people here suggest?

View on HN

So how do you "launch" a beta, what do you to do interest the relevant startup blogs and generate some publicity and attention for your freshly baked business?

I don't know why you're targeting startup blogs unless you are intending on selling your services to startups (which has its own problems, chiefly that a million other people are chasing after the same $600 dollars in their bank account). Michael Arrington will have, to a close approximation, the entire population of Silicon Valley send him an email today. Your odds of getting even fleeting attraction are low. Should you get fleeting attraction, you'll find your site crushed under a wave of looky-loos who will make disparaging comments on everything they can before abandoning it.

By comparison, the people you made your application for might actually, you know, want to use it. Why don't you target their blogs? Speak to the people you know in their community? (You do know people in their community, right? If not, turn off the IDE and go meet some people in their community.)

I'd also start your SEO efforts, if you haven't already. (If you haven't, turn off the IDE and...) That is a deep subject when starting from nothing, but the short version is think of ways that linking to you will improve the lives of people who don't want to buy your software, then, tell them that.

Oh, and this comment is going to sound like a broken record from me, but launch is not a one-time event. 99.9999999% of people who see your site will not see it on launch day. For someone who comes in on day 4 or day 40 or day 4,000, your site JUST launched. Optimize for their experience, not Michael Arrington's.

For mine I just asked on the relevant subreddits to come check it out and give feedback. Got my first 100 users that way. + they helped me design the site: http://styleguidance.com

As far as bloggers, unless you invented Google(or worked there) or have millions in funding, it's probably not worth bothering since you won't get much coverage(from the bigger sites). And the type you'll get will most likely get you a few hundred visitors. i.e. a 50,000/mo blog, sent me 5 people.

And the traffic you get from doing a "launch" is usually garbage anyways. And will evaporate within a few days. Unless of course your core users are entrepeneurs who read tech blogs. Then you might have some luck.

Basically a launch doesn't really get you anywhere, you should see "launch" days, as a way for you to declare for yourself that you are open for business. So that you can get serious about marketing.

i.e. for my launch date, we got 5 stories, and a total of 668 visits. That was November 18th, and last month we had 35K visits, Page Rank 5, Sub 100K Alexa Rank, Sub 20K U.S. Alexa Rank, 2K questions(content), 1550 twitter followers, 960 Facebook Fans.

So don't worry about planning for a spectacular launch, just do it, and then get to work.

I wouldn't mind reading about how you "got to work" after the launch and managed to build that traffic in about a month. Care to go into more detail, either here or on your blog?

mostly promotions and SEO.

I'll see if I can write something up later, since it takes a while

I wrote down a few tips I'd collected a while back: http://david.weebly.com/1/post/2008/02/press-for-startups-10...

The landscape may have changed on the embargo bit (TechCrunch claims not to respect them), or not -- they still seem to respect ours and most startups I know.

do what you did here, but provide a link;)

Ideas:

1. put beta in your logo and make submit feedback forms as prominent as possible without being annoying

2. maybe include a countdown in days until official launch on the site. reinforce in peoples minds that they will want to come back later or keep an eye on the site.

3. Since you want it to be a 'beta launch', just launch it now. Start pulling users in, they will be able to help you make a bigger media bomb for when you decide to 'launch launch'. hit up all the aggregators and even if you dont get a big impact you can hit them all up again when you decide its no longer beta with a 'launch' story

I feel like you should hand select your first 25-50 users, a user base that size will give you plenty of feedback.

Once you've done that, I suppose it's time to get more serious about acquiring users.

Yep, feedback would be the key for me.

Yeah we've been 9 months in development and ~6 months in private beta, got a few dozen users who've really helped shape the direction of the system and turned it into something amazing.

It's really the "what happens next" bit that I'm not sure about.

I'm not a fan of stealth or private beta. I think the best plan is to not have a beta launch.... just start organically collecting users.

No one of significance will notice your software sucks, but when your software is good, investors, customers, and partners will notice.

It hasn't really been "stealth", it's been "controlled", it's an analytics service (not for websites) and we wanted to be sure it worked + not get raped by traffic we couldn't afford to scale for during development.

I think your fears are unwarranted. Every entrepreneur seems to have these happy fears. Oh no, what if my site crashes under terrific load! What if super-VC Paul Schmaltz sees a bug in my website!

Even if you get enough traffic to knock down your server, that will just be a boon.

And it's better to get your software out there half-working and get feedback than to build something nice and have to rebuild it after you find out the average user wants something else.

re: launching

I soft launched http://newsley.com in November via comment here on Hacker news. I wasn't interested in a big splash, but I was really interested in getting a couple of beta users that could give me some good feedback, and that's what I got. It's a good crowd to get some feedback on what's broken and what works.

Since then, I've posted a few more links here in the comments, and prakash posted a link in one of his comments on a thread regarding economic social news sites (thanks). I've also posted a link asking for beta users on Craigslist San Jose, and I've posted a couple links in economic blog comments where appropriate.

Traffic stats so far for 2 1/2 months: 1200 uniques, and close to 5000 page views.

My future marketing plans include buying some more beta users off of Mechanical Turk at $0.05 a pop, and using a $100 credit I have with Adsense. I have a hunch, I'm going to get a lot better traffic with Mechanical Turk than Adsense.

re: funding

The scuttle butt in the Valley is that funding is slowly starting to flow again, but it's still pretty hard to come by unless you have traction. Traction == customers signing up and loving your product and possibly revenue.

I'd focus on making your users really really happy, and slowly build up a customer base from that rather than go for a big splash and hope that things take off.

There are some great comments here. I don't need to repeat what's been said, but would like to add that you should not go for a big launch yet. Instead, invite some people who fit the definition of your core audience and ask them to test the site out. Use services like uservoice.com and survey.io to collect user feedback and make sure your product has a solid market fit. Doing this will make sure you don't miss the mark at launch, while at the same time providing some key insight for messaging to be used with a more prominent launch. Good luck!

I think you should ask yourself if you really need funding in the first place.

The only way we could go it alone would be with no free trial, but that's a pretty tasteless option since we're selling something that's never really existed before to a market that's going to want to play with it before they pay for it.

It's an analytics service for casual / social games and in some cases we're talking games that get millions of plays (we've got several like that in our system already) which can translate into tens of millions of events being logged, and while we've spent tons of time making everything as streamlined and cost-efficient as we can there are unavoidable expenses that might predate our ability to pay for them by ourselves.

Hi Ben,

I appreciate your concern about scaling your service but I think you should consider a fremium business model that allows you to accommodate more free users at lower cost.

Instead of logging tens of millions of events for free users, you could log just a subset, which will allow you to use far fewer servers whilst allowing free users to "evaluate the usefulness of your new service".

It will also make sure that the future of your startup is not wholly dependent on external investors.

We've done something similar with Trafficspaces. We run our service off Amazon's cloud system, which gives us a lot of flexibility.

If you need assistance on how to structure your freemium model to make sense, send me an email at niyi [at] trafficspaces [dot] com.

Reply on news.ycombinator.com