Search This Blog


Thursday, October 11, 2007

Five Things to Remember with Site Development

I often find that as we continue to move at an incredible pace during site development we become too involved with our end product and lose the user perspective that is so important to the site’s success.

To mitigate this risk try the five following tactics:

1. Take a step back and look at your site/section/page, ask yourself, what is the message you are trying to commutate? What is it that you want the consumer to do? Is this clearly laid out that anyone looking at the site would give the same answer as you?

2. Test, test, test and this doesn’t mean slow down. When you have a new release or change, ask people that are not attached to the development what they think. This could be as advanced as a focus group or as simple as walking around the office and pulling up the change on people’s computers.

3. A/B test, see how the change is adopted before making a mass change. You don’t need to make investments in software to support this (it just makes it really easy) as it can be as simple as trying out the change for a period and then reverting back to see how it impacts the goals of the project.

4. Understand what a success is and whether the change moved you in that direction. If not, you may have missed the target or you don’t understand what the levers really are.

5. Watch the metrics – a key one to watch is bounce rate, are consumers interacting more with the site/section/page or are they turning around and running for the hills - more than before. If you have an analytics package (and everyone should as Google Analytics is free) look at pathing, flow and fallout to understand how the change effects what the customer is ultimately doing on the site.

To be successful you should build this in as a common practice not just when you do something big. Many of the above can be done on a regular basis and having this understanding can only help you more in the future. The one common key stream in the above is the customer. Listen to what they are saying and build a product that they want, not one you think will work for them.

No comments: