Releasing Too Early, Reprise

A few weeks ago, I wrote a post about Steven Johnson's Outside.in geolocal social media app. In particular, I complained about his (and many, many others) decision to not implement some much-needed social features in the first release:

I am glad to hear that the folks behind Outside.in are working hard on building out the social dimension. But I disagree on his take on the chicken and egg problem of new apps: you shouldn't release apps that are social without the Social Tipping Point that will allow social growth to occur. If you leave out things like profiles, discovery of likeminded people, and so on, you aren't really priming the pump, because people cannot connect. You may have alerted folks to the existence of the site/service/product, but you will be positioning it badly, since the user experience is socially poor.

After you have defined and built the Social Tipping Point for your app, then release, and continue to release often. If however you release before the STP, all you have done is released is some other app, one that desperately needs a social redesign. In this case, you can lose the momentum associated with initial release, and are in effect dooming your site/service/product to a relaunch later on.

In this specific case, I think Johnson and his partners should have launched with both the prepoulated content and the STP of features, and not until both were ready.

I have been through this scenario with several clients in 2006, and in all case, I regret to say, my concerns were justified by subsequent events. In all cases, social apps released before the STP is completed had initial surge in interest followed by a rapid decline, and then a stall.

Well, the same contorted logic has come around again, in the recent release of Buzzfever, which I bitched about on 3 Jan. Curtis Summers of Buzzfever left a comment and a trackback to my post, where he basically goes through the same litany, simultaneously agreeing that the app needs the features, and that he is going to provide them later on:

[from Missing Some Features... You Bet]

Stowe Boyd has a great post about BuzzFever. As you read through it, with phrases suggesting BuzzFever needs a “social facelift and tummy tuck. Otherwise it’s dead”, you may not think that’s so great. Allow me to tell you why it is: he’s exactly right–BuzzFever needs more social connection.

Here’s the better news: Every feature he mentions is already on my TODO list for version 2.0. In fact, I actually dropped some of the social and grouping functionality off the 1.0 feature list just to get BuzzFever out the door sooner.

I initially lamented dropping features off of my ideal vision, but now I’m glad I did. Since launch, I have a renewed motivation to tweak the site and push forward more quickly on 2.0 features. I’m the sole developer for the site, so having motivation to keep moving forward is extremely important. Launching renewed my coding motivation.

So, BuzzFever IS lacking some social features now, but rest assured that these changes are on the way.

So, again: Why release a product lacking critical features? You are dooming yourself to a relaunch of the later, more complete product. People's initial response to the incomplete product could damage your brand and positioning. It's nuts.

But people continue to convince themselves to release too early, "to get something out there," no matter what. It's like serving the first half of a meal when its undercooked, just to get food in front of your guests. But its inedible!

I was chatting with Tony Conrad of Sphere about this yesterday, and he reminded me of a post he made early in 2006, when he decided to delay release of Sphere in order to get new, critical features in the initial release. So at least some entrepreneurs are willing to counter the now conventional, and flawed, thinking around releasing too early:

[from c’mon already!]

End of December [2005], we made a decision to add some very cool, advanced features to the site that we believe pushes the quality of blog discovery to a very different level. Delaying our launch was a very hard decision, one in which we went back and forth on for weeks. Ultimately, despite the great beta feedback, we felt it important to take a step back and work towards our larger goal. I’ve been involved with a number of start ups (Oddpost; Post Communications; Iconoculture; and Automattic/ Wordpress - by the way, this is my first post on the wordpress platform - very nice!) and I believe you need to trust your instincts. In this case, our instincts tell us we can do something really great so we’re focused on getting there. Once we launch (soon), it gets harder to develop, especially with a small team (we’re 4 people). So, we’re in that mode right now, hyper focused, making real progress at a fun rate.

I expect that I will get to howl about this at least 25 more times in 2007. It's a lesson that new start-ups seem congenitally unable to learn, until they have burned their fingers in the fire.

tags: releasing+too+early, social+tipping+point, curtis+summers, buzzfever, steven+johnson, outside+in, tony+conrad, sphere

Comments

[Ted Rheingold]

There should be a word for the first release sweet-spot. That point when you have enough core features that the purpose of a service can be understood and used from day 1, but the bells and whistles are not there.

I also think when releasing a new project you should not focus on what will be launched in version two. You should release features one by one as fast as you can. Just keep adding them. In fact the user feedback you get will likely shred any pre-release ideas you had on what version 2 will be anyway. Just get in complete release mode and push out everything you can and reprioritize to do lists on a weekly if not daily basis

[Curtis Summers]

Stowe,

I take your points, but I'll argue that the motivation gained from launch for me personally was necessary. I'm not a team of four, so internal motivation is admittedly difficult to maintain.

> Why release a product lacking critical features?

You see these as critical. I see these as important, but not critical. I think BuzzFever's present set of features stands on its own and makes for an interesting site. If I didn't think so, I would not have launched.

I'll leave it at that for now. Only time will tell, right?

And, if you're correct, I'll bow before your Web 2.0 wisdom!

Curtis

[Steven Johnson]

I'll second what Curtis says:

"You see these as critical. I see these as important, but not critical."

For outside.in, our primary reason for being is not social networking. We're trying to organize geographically all the disparate information out there related to local communities. Social networking will certainly be a part of that, but the features we launched with were in fact more important to our ultimate mission -- and more unique -- than social networking features.

If you look through all the responses to outside.in when we launched, you'll note 1) that the response was overwhelmingly positive, and 2) that by far and away the primary complaint was not the lack of social networking tools, it was that we didn't have enough content yet. The consensus seemed to be: the tools are great, just needs more data.

Don't get me wrong -- as I said a few weeks ago here, we definitely want the platform to enable neighbors to meet each other. But there are other more important functions that we wanted to lead with, in part because there are a thousand social networking sites out there, but very little out there dealing with the geographic web the way we are...

[Curtis Summers]

Ted,

You bring up a good point that reflects what I'm doing with BuzzFever. Stowe's feedback and the comments of others have helped filter down what new features to include first. I have a huge list of "maybe" features, but the feedback I've gotten in the past couple of weeks has been invaluable in narrowing my scope to the most wanted features.

And, I'll add that this feedback is, I think, from a broader spectrum of folks than would have signed up for a private beta.

[Stowe Boyd]

Steven - The point that I am making is that the social features are necessary for viral adoption, not as networking for its own sake. And I agree that lack of content is an issue, a critical one. I am simply recommending that products shouldn't be released until both needs are met. I know that we disagree, but I still maintain that viewpoint.

[Dan Keldsen]

Stowe (and all) - Just caught this entry. Deciding which features to lead with is definitely tricky, and of course everyone has their own opinion. I definitely follow your thinking however - it's the viral aspect that would help lead adoption/spread, and likely also feed the need for feedback on the part of Curtis. At least that's how I would think as well.

New startups are (thank god) following a more or less Agile method for development these days, but being on the constant improvement gerbil wheel does mean that the most important features to not just having a working prototype, but a virally intoxicating prototype is certainly key.

Even the mighty Google has found that while Google video did and has made some nice waves, that upstart startup YouTube blew right past them, largely because of the features Stowe is talking about.

At least with an agile method of working however, it's possible to iterate quickly back on track - versus my typical enterprise clients who iterate on a 3-5 year timeframe at the fastest (somewhat overblown, but not much really).

Cheers,
Dan