Ecommerce website analysis – what we discovered analysing over 200 ecommerce websites
Ecommerce website analysis or heuristic evaluation is the starting point of the Worship conversion rate optimisation process.
Over the past few years, we’ve analysed hundreds of ecommerce websites. In that time, broadly speaking, we’ve seen 3 categories of website.
Website Analysis Category 1 – ‘The Good’
Ecommerce websites that consider the user and what they need at every point in the conversion journey. They’ve taken the time to research what’s important to their users, what users do on their website and where they can plug their sales leaks.
They’ve been using continual A/B testing to incrementally improve for years. They see year on year growth and consistently high conversion rates. These websites aren’t necessarily the most beautiful websites (think Amazon) but they are well designed and they make buying as easy and painless as possible.
Website Analysis Category 2 – ‘The Bad’
Ecommerce websites built on best practice principles but never analysed or tested to facilitate continual data-driven improvements. They’re managing to grow but only because they’re spending more and more on traffic each year.
The websites look pretty but they are more concerned about the way their brand is presented than the user experience. The result is whole host of issues such as:
- Important content being pushed way down the page
- Typefaces that are hard to read
- Important content being hidden to make pages ‘cleaner’
- The website search being almost invisible to ensure the logo is the most prominent element in the header
- The mini basket in the header being presented as a very small subtle icon to ensure it fits with the brand ethos
- Checkout fields so faint it’s hard to see where they start and end
Often these sorts of websites have to go through 3 departments to get changes ‘approved’ and changes are pushed live without first being A/B tested. They can’t implement an A/B testing programme because their website is governed by brand guidelines and beauty.
Website Analysis Category 3 – ‘The Ugly’
Ecommerce websites offering a woefully bad experience, visual design from the late 90’s with dwindling traffic to boot. They’re too set in their ways to embrace the tools and techniques which would facilitate continual improvement.
In a few years, they will pay a digital agency £100k to redesign and rebuild their website and none of their analytics data will be used to inform the design decisions because the agency won’t want to be bound by the current website, even if some of it is working well.
A few years after the redesign, they won’t exist and they’ll blame market forces/Amazon for their demise.
First impressions when analysing an ecommerce website
What’s interesting is that when we first arrive at an ecommerce website we’re going to analyse, it’s not always immediately clear which category the site falls into.
Some of the most beautiful sites fall into the last category and some of the sites that, at first glance seem slightly dated, fall into the first category.
An example of a website on the brand beauty trackWe’re not saying that your website can look like it’s come straight out of 2002 and still be considered good, it can’t. But the point is, beauty is just one component.
Many people seem to prefer their website to meet their personal taste than convert visitors into sales. So design decisions are based on personal taste and gut instinct. This presents a huge barrier to growth and leads to a lot of wasted time and energy.
Many people seem to prefer their website to meet their personal taste than convert visitors into sales. Tweet This!
Which website analysis category are you in?
If you want to know which category you fall into, ask yourself the following questions.
How much do we really know about our users? Not broad demographics like age and gender but answers to important questions such as:
- What motivates our users to buy from us?
- What do our users do with the products they buy from us?
- What’s important to our users when buying from us?
- What stops our users buying from us?
- Where else do our users look for our products?
- Where do users get stuck on our website and why?
- Do users understand our core value proposition?
- What research do we do to discover what people are doing on our website?
- Where are the biggest drop offs on our website’s conversion journey?
- Are there any patterns of behaviour for groups of users that convert at a higher or lower rate than the website average?
- When was the last time we used data-driven insight to improve our website?
If you can confidently answer these questions because you’ve done the work, there’s a good chance you fall into the Good category. Well done, keep up the good work.
If you haven’t got good answers to these questions then there’s a good chance you fall into the Bad or the Ugly category.
Changing from a Bad or an Ugly to a Good
The good news is, you have the power to change your category. And it doesn’t mean rebuilding your website from scratch (probably).
You can start discovering what your users think, what they do and how you can improve today by:
- analysing your Google Analytics data
- using a heatmap/scrollmap tool
- running surveys and polls
- and doing user testing.
All of this can be done fairly cheaply, but it does take time. You also need to ensure you’re asking the right questions. We’ve seen ‘Do you like our website?’ in surveys more times than we care to remember and it’s a pointless vanity question.
If 87% of people say they do indeed ‘like your website’, then what? Everyone cheers and goes home for the day, happy in the knowledge that the website is a good as it can be? Your conversion rate is still 2% right? So there’s still work to be done.
If you would like help with this process, please contact us on 0161 713 2434 or [email protected]. Our conversion research is extremely comprehensive and results in actionable insights and testable hypotheses. Do your business a favour and become a Good today.
22nd August 2016

Get a free 15 minute CRO strategy call with a senior optimiser
Request a callback