Don't treat your website like a commodity
The traditional approach to product development involves coming up with new idea and then driving as many people towards that product as possible, in the hope that some of them will want it. As such we adopt the language of marketing, and talk about marketing funnels and conversion rates.
If our marketing department has done a good job they will have created a campaign that not only generates traffic, but creates a previously unrecognised need. Tired? Need a break? Why not have a KitKat?
This approach treats every product like a commodity to be bought and sold. However the problem with commodity marketing is the fact that there is very little difference between products. There is so little difference between a Snickers and a Mars bar you can't really compete on features. And because commodities are generally cheap you can't really complete on price or quality either. So instead you create brands, build tribes and keep pumping money into your marketing campaign to ensure that when you're standing at the chocolate counter with your 50p, for some reason your brain says "how about Twix" instead of a hundred other confectionaries. It's not very subtle and is a bit of an arms race, but in the word of commodity marketing it's really all the ammunition you've got.
Sadly most website owners treat their online products in a very similar way. Rather than creating something that users actually want, they'll simply extend another product or service. So just in the same way that Mr Snickers may have looked at Mr Mars and said, let's make something similar, but with nuts, your average web entrepreneur will take an existing product, service or idea and extend it slightly. Even if it's a genuine innovation it's usually the idea that comes first, rather than the need.
Because there is little difference between the majority of online products, product managers resort to old school commodity marketing and focus their attention of driving traffic to their site in the hope that a few people will stick. This is why most big online products seem to focus more of their attention on SEO and marketing than they do on the product itself. After all, in their eyes the site is just another product to be bought and sold. It doesn't really matter if it's actually any good.
To get around this problem, people need to put a lot more thought into projects at their inception. It's easy to fixate on a good idea, but our first ideas aren't always then best. There's nothing to say that there aren't even better ideas out there waiting to be discovered. So rather than jumping in there with the first idea that makes the grade, try to analyse the problem from the users perspective. Does what you're building solve a real or simply an imagined need. Are there more important, yet to be discovered needs that your product or service could solve instead?
Go out there and do some research. Watch people use existing products or services, both online and off. Ask people what they like about the services and what they dislike. Try to find out what frustrates them about the current status quo and use this to look for unmet needs. Rather than creating just another "me too" product, see how your site fits into the bigger ecology. Rather than compete with the market leader, try to compliment their services and fill a gap they're not able to meet. Create prototypes and test them on real users. However don't just do this to see if the site is usable, do it to see if the site meets users needs and be willing to ditch the prototype or even the project if it fails to test well.
I think one of the reasons most online products fail is through a lack of research, planning and customer empathy. We put all our eggs in one basket, without knowing if it's the right basket or not. We end up committing to one course of action far too early in the process, so by the time the signals start coming in it's too late to alter course. We have this fear that the time we spend on planning is wasted as it could be spent on production instead. However research and planning can not only help you avoid costly mistakes later on in the process, it can ultimately help you build better, more fulfilling products. Products that solve real world problems and make you more money in the process.