Stop Adding Features
Seriously stop it. I had read all the books about launching and running lean startups. But somehow, I still managed to constantly add new features to Task.fm. So whats wrong with new features? With each new feature, you add an extra layer of completxity. This is an extra layer that your customers need to figure out how to use. An extra level that needs customer support. An extra layer that adds to server load.
Basically, if you're not careful you'll end up with a big mess. I'm still recovering from this myself - and yes I've learnt my lesson. And this leads to my next point...
Don't Promise What You Can't Deliver
Start running a web app, and sooner or later you'll start receiving emails asking for various features to be added. Its tempting to say "cool idea. Yes, we'll look at building that in the future". or "yes thats on our todo list". But, you'll soon discover, as I did, that in between announcing you'll build a feature and the time it actually takes to ship it, a lot can happen.
Thats why I've sort of fallen out of favour with GetSatisfaction. It allows users to post and vote on feature requests. But remember, your users don't understand the complexity of a new feature and how it will effect their user experience. I've shipped features that had received a lot of requests. But after a couple of days of using the feature, the "thats cool" effect had worn off and they stopped using it. So when an email starts with "wouldn't it be cool if", just reply with a generic "Thanks for the suggestion" type email.
Another mistake I made was not providing the proper level of communication when an app went down. Cloud based software will always have downtime. Its inevitable and often may be out of your hands. Don't be tempted to bury your head in the sand and hope it will just go away.
Communicate on Twitter, your blog, via email - however you can. Make sure your users know whats going on.
If you think you have a better product, don't be tempted to cut the price to match your competitors. If your product is premium, charge a premium for it. Remember, your app, unless its a straight clone, can never really be compared to any other product. It can always have unique selling points that you can justify charging more for them. Your users will know this to. Its often amazing what people will pay for - you think a customer is paying for your cool features, but have you ever considered they chose your product over a competitors because they like your sites colour scheme more?
Ignore the Critics
Sometimes people will write bad things about you or your product. Its often the hardest thing in the world to not respond, but at the end of the day, its the best thing to do. I'm not saying block everything out, just don't feel the need to go out on a PR frenzy and respond to every comment thats made about you. Take Digg.com for example and their latest product flop. If they stopped feeding the big blogs with constant "we screwed up" lines, the blogs would have stopped writing negative posts and moved on. You don't need to justify every decision you make. Sometimes you're allowed to say "we think this is better".
There are my 5. Tweet me up with some of yours.