There are so many considerations for UX practitoners when building a site design that it’s been tough for us to come up with our top 3. Nevertheless, here they are.
1 – Consider accessibility first
We get accessible design. There are a host of good reasons why designs should be accessible to disabled visitors and, in some countries, strict laws which enforce accessibility as a legal requirement.
However, accessible sites requires a different design mindset to rich-media, interactive designs. This is because accessible designs hinge primarily on clean code and ordered content, rather than visually engaging themes or interactive elements.
During planning, consider the question of accessibility first. If the site should be accessible, follow the principles of progressive enhancement and build a basic, clean design which can be enhanced in non-obtrusive ways, whilst respecting user preferences.
2 – Involve target audiences
A truly usable design is customised to its audiences. That customisation should be based on facts about those audiences.
If you are lucky enough to be working for an organisation which has already defined its target audiences in their communication strategy, don’t assume that their assumptions about these audiences are correct. Do some research of your own with focus groups composed of members from your customer’s target audiences and learn about them from the horse’s mouth.
Many ineffective sites are based on assumptions about target audiences, rather than the hard evidence of interviews and focus groups. Such measures are sometimes considered unnecessary or expensive, but stick to your guns and explain to your customer why they are the key to understanding users and their needs.
3 – Test, test and test again
Testing measures the success of a design and there are many rounds of testing that must be undertaken.
systems testing engineers will run functional and system-level tests, infrastructure engineers will run load and balance tests on their web servers and you will need to conduct usability and acceptance tests with users.
Conduct these tests directly with your focus groups. Get them to use the design and observe them while they do it, preferably using non-intrusive techniques such as video cameras or software like Silverback rather than some of the more traditional approaches, such as one-way mirrors, which we find place users outside their comfort zone and make them feel ‘observed’ which, in turn, makes them less likely to operate normally.
We also find it better to run tests with users in groups rather than individually, as your users will be less self-conscious if they can confer with their colleagues. We find this generates better quality data from which you can draw your conclusions.
Finally, test your designs regularly and especially during early development. Early testing helps to identify potential issues before designs become ‘baked’.