website analytics and testing.

Data is perhaps the most important part of reviewing a website’s performance and it’s important that where possible, data is measured against your online marketing model and goals. This gives you a read of how you website is performing based on how you would like it to be and can often (with the right analysis) point us in a direction of what we can do to help better attain your goals.

I can create monthly reports analysing trends, pinch-points and popular content as well as comparing them to the previous months to measure whether any design changes are having an impact on the site.

How do you measure a users experience? Do we sit down users and watch them use the website in a lab? or is there a way we can view how our website’s being used by monitoring our visitors usage of the site.

I’ve recently been experimenting recording user sessions on some client sites to ensure that some recent design changes have been well received. The sessions are recorded remotely using a script that runs from the site. It’s very lightweight and doesn’t affect load times so the user doesn’t experience anything from the norm.

A particular useful tidbit of information was gathered on a client site recently when I noticed that users were highlighting the product name and model number on the product page – we’ve all done it right? Paste it into Google and see if it’s cheaper elsewhere. This nugget of information allowed us to create a micro-interaction that told the customer that they would refund the difference if the product was found cheaper 30-day’s after purchase and as a result, sales went up.

As the old adage goes ‘there’s more than one way to skin a cat’ – though I don’t imagine many cats get skinned nowadays, it’s still a widely known and used phrase and bizarrely, has a place within web design.

All websites have a purpose – website owners want their visitors to do something. Thus, for instance, if you have a blog, you want your visitors to engage with your content. If you, on the other hand, have an e-commerce website, you want your visitors to buy as many products as possible. And if you are offering a service? Then you probably want your visitors to contact you and ask for a quote. The more people do what you want, the better for your business right?

So, back to the saying. There’s likely a very efficient method of skinning a cat that works 80% of the time and one that’s only works 40% of the time. A/B testing allows different pages, content or styles to be shown the users at differing times to better understand a more effective way of selling. One page that’s laid out differently with slightly less content may generate more leads than another page that’s had lots of input from different figureheads within the organisation and looks a bit overwhelming (we’ve all been there, right?)

We can then use this data and these experiments to present better performing pages. Maximising set goals for the site or page.