This week the Future of Web Apps (FOWA) conference was held in London. Unfortunately, due to a bad landlord, I was unable to make it over to the conference till the end of the last day, but nonetheless managed to catch a very interesting panel on some of the lessons that successful web application developers and companies had learnt over the last few years. I also had a chance to get a feel for how the industry is feeling in the wake of the sub-prime crisis and a general economic slowdown. I had anticipated that the tech crowd might be subdued; even concerned, by the threat of a cut in investment, but it seems that the community is still financially bouyant, and there is no shortage of interesting new applications and programming languages to keep innovation happening. (I particularly like the look of the new Adobe Air system for taking web based applications onto the desktop).

Some snippets, then, from the panel discussion. I made very terse notes, but there were a few moments of enlightmenment; most notably the founder of Dogster/Catster (social networks for pet owners...), whose epiphany came when he realised that one could not “just go online and look at pictures of dogs”. Right.

There seemed to be a lot of surprise from these entrepreneurs and developers at just how much uptake they had had for their products; often in the initial phases, while they were still testing their products, and the difficulties this could cause from too much customer feedback too early. the lesson? Remember that the customers who email you most are not representative of the rest, and make sure that you roll out new features for your standard users, not the really advanced ones; you can always extend them later.

Recognition was especially due to a couple of these sites: Ficlet and Satisfaction.

Ficlet enables user generated fiction; people can write prequels or sequels towork already on the site, allowing the overall body of work to grow in a controlled fashion. Even more excitingly, it’s the first AOL project to use Creative Commons licencing, and it’s at the vanguard of AOL’s adoption of openID.

Satisfaction is a site that enables companies to use their customers to provide each other customer service (at it’s simplest!): it allows questions to be asked around brands and products, and for either the company or other customers to answer them. Answers are persistent, too, so if someone has asked your question previously, the solution is right there. It also has a lovely user interface and a lot of hip web companies using it’s services. Definitely one to watch.

The final couple of items on the panel were about make-or-break moments in the companies; what mistakes almost brought them down, and what has ‘just worked’ for them. Some of the key answers?

Mistakes

• Not hiring enough managers
• Being invite only when they launched
• Being afraid causes wrong decisions. ‘Just do it!’.
• Stick to the launch date...don’t get caught up in details/business side too early
• Get customers and a product before trying to sell
• Thinking you need to make a decision generally means you will make that decision wrong.

• Make your product as viral as possible.
• Start small and try and stay small as long as possible.
• Figure out the hook that your application needs. Small solution to big problem.
• Using Ruby on Rails (a web programming language) and be Agile.
• Waiting to make a decision that needs to be made (esp recruitment)
• Get great partners.