starting to be apparent coming out of some of the spectacular dot-com failures? Soundbitten’s article on the failure of Verde and the possible contribution Scient made to the failure is instructive. Verde was trying to be a web-only content site, with e-comm built in at the ground level. OK, fine. But they outsourced the very lifeblood of the company – the platform on/in which it was to live!

Maybe it’s just me, and my biases (the company I work for does its own design, programming, editorial, hosting – everything) but I think that if you’re going to live on the web, you have to develop for the web in-house. That’s the real challenge for marketing and sales types with an idea – to figure out how to work with developers – and developers who themselves are radically different from one another (i.e., programmers and designers are different from one another in dozens of ways, in general). But if you’re a retail bookseller, you don’t farm out your retail sales staff to a consultancy – that’s what you’re about, to a great extent.

There may be a place for consultants in all of this, don’t get me wrong. But if you’re a net company, you have to have to develop your own internet infrastructure – you can’t get around it. Doesn’t mean you can’t purchase products that will help you do this – a company doesn’t have to invent everything from the ground up.

Maybe the real lesson is that new dot-coms try to separate back end from front end too much. It isn’t enough to be good marketers, writers, strategists. To split form (content or marketing) from function (CMS, design, UI) is to tie one hand behind your back as you try to get a dot-com off the ground.