Design is and always will be subjective, whether it's brands, fashion, interiors, cars or whatever. Opinions differ wildly from person to person on what constitutes a beautiful room, gorgeous car or easy-to-use device.

So, as web designers, how do we get beyond second guessing people's taste towards something a bit more objective? A situation that occasionally arises when working on web design projects is that a client's taste (or a designer's) starts to affect the choices that are made. In reality, we all put a little of ourselves into a design – which is in fact essential to build in personality and to make it "human" – but what if opinions simply differ and the project stakeholders have different taste; should that button be a subtle shade of blue that fits with the brand or a bright red to purposely stick out like a sore thumb and demand the user's attention?

Who wins the argument? The professional designer who's got years of experience of what works? The person paying the bill? The most senior person in the room? The loudest person in the room? Someone's mate who's quite handy in Photoshop?

In reality, I've experienced projects where all of the above have had significant impact on design decisions, often to the detriment and success of the website. And of course, if all of the people in the room have different opinions (even changing opinions during the design process), we can end up with a reactive situation where we're trying to please everyone by creating iteration after iteration after iteration... version 27 of the homepage.

If you boil things back down to simple objectives of pretty much every web design project, you'll see that the word "user" crops up again and again:

  • We want the user to spend longer on the site
  • We want the user to explore more pages
  • We want the user to spend more
  • We want the user to share more
  • We want the user to come back more often

The user, the user, the user... who is this person and if they're so important how come he or she doesn't get to input into our design process? The user is, of course, anyone who might visit our website and the only way we can find out what they think is by asking them, or to be more precise, by asking a representative group of them.

But user groups are difficult to organise, expensive and often inconclusive, I hear you say. They can be, yes and the problems are perhaps best explained by the infamous story of a user group that Sony ran to ask the public what they thought about a new yellow radio they were thinking of launching versus a more traditional black option. Everyone in the user group stated that they preferred the yellow one, but when they were allowed to choose a radio to take home as payment for their time, every person in the room walked out with a black one.

Thankfully, there's far easier and more reliable ways of testing websites with users. A really efficient method that we employ is an online service called For a small fee, they will recruit users based on our specified criteria, such as gender, age, location, income bracket etc plus any other qualifying question we want to throw at them. As few as five users can give valuable insight. Once recruited, they are then sent a link to the website to be tested and a list of instructions that we've written, telling them what we want them to do on the site. This can be open-ended, such as "Explore the site and tell us what you're thinking as you do so" or a specific task, such as "Using the site's main navigation, find a pair of trainers that you genuinely like and go through the process of buying them".

The system captures everything that happens on the user's screen and records their voice, so that we can see where they're clicking, how they're progressing (or failing to progress) and hear what they're thinking at every point in their journey. Tests usually last around 15 mins, which is ample to give us great insight into how our site design is performing and where improvements can be made.

Watching the user videos is always enlightening, often surprising and sometimes even entertaining. For a small outlay, we have gleaned huge amounts of insight on where a website is working, where it isn't and what we should do to improve it. Gone is the subjectivity, we now have real people using the site in as real a situation as we can hope to achieve and telling us what they think.

If that on-brand, subtle blue button gets overlooked by every single user and prevents them from finding their way through the checkout, all of a sudden the bright red one starts to sound like the better option. We've replaced opinion with fact, tested with something more tangible. Of course, you can get different results from user to user and it's important to learn how to spot the patterns, not just react to every little comment that might be made.

There are all sorts of ways of carrying out user tests, but as a quick, inexpensive option, or any other similar online service is worth its weight in gold. If you want to find out how your website is really performing and how to make it better, I highly recommend it.