Why you shouldn’t launch your startup in the press

The bottom line is: manage hype, and don't buy into your own. One of the most interesting takeaways I got from hearing Eric Ries speak a few weeks ago, is how toxic hype can be to startups not just in the press, but also among investors, friends, family, etc. As entrepreneurs we face myriad challenges and problems to be overcome between vision and product execution, and these massive expectations that we build up can be a huge barrier when it comes time to make tough decisions and revisit product strategy, especially when (not if) things don't go as well as we would like the first time we deliver products into customers' hands.

Tried this with a previous startup. Soft-launched an iterated on the idea. A few months later, started pitching the idea to press. The reaction we got was very negative—no-one wanted to cover us unless there was "news". We had already launched so there wasn't anything "new" to report. We had to invent features to put in the product and sell that "story" (it worked, sort of: people wrote about us but reported on the new feature rather than the fact that we existed).

I recommend putting some sort of beta-wall or limited access on your launch if you're going to do this, if only for the excuse of being able to announce public access later. I wish PR tricks like this weren't necessary, but they really are.

I can’t speak to recent events, but a decade or so ago when I was involved in some venture-backed startups, there was a lot of pressure from our financiers to launch in the press, while we wanted to launch (and possibly fail fast) in private.

As you can imagine, sometimes the interests of the entrepreneur and the interests of the VC are not aligned. They benefitted from press about their investments, while we benefitted from organic growth. They felt that even if this cost us revenues, it would be made up for by subsequent investment rounds at high valuations driven by our press releases.

n=1 and all that, perhaps things are different today than in 2002.

The thing though is that I think the marketing launch shouldn't be done as late as possible. Rather it should be done once you have a strong understanding of what your value proposition is. This takes time and it takes listening to customers, figuring out what works for them and what doesn't.

This means yes, do the product launch as early as possible, and if you are moving into an established area, market early too, but if it is a new area, be more cautious and put off the marketing launch until you have typical customers, and know exactly what you are doing for them vis a vis the competition.

I don't think it's fair to lump Wolfram Alpha in with Cuil and Color. I use Wolfram Alpha at least once a week. It's great for sharing plots of arbitrary equations and running through back of the envelope calculations.

Cuil immediately came to mind after reading this: http://www.nytimes.com/2008/07/28/technology/28cool.html