Heuristic analysis: how to evaluate a website
What is a heuristic analysis?
During a heuristic analysis, UX experts identify areas where a website could be leaking profits. Although it’s a subjective research method, it’s carried out by experienced optimisation professionals. When approached with an open and counter-bias mind set, it can be a valuable asset.
The evaluation usually has two objectives:
- To identify “areas of interest” where the optimisers believe the website isn’t providing a good enough user experience or where it could be leaking sales. For example, the site lacks a clear value proposition on all landing pages or the navigation is confusing.
- To pose specific questions in order to formulate the priorities and approach to the optimisation project. For example, the optimiser considers the level the website is at and decides the appropriate course of action.
The output from both of these needs to be validated with actual data (from quantitative and qualitative user research methods).
A heuristic evaluation is as close as you’ll ever get to “best practices” or opinions in CRO. For instance, having big buttons that can be easily tapped on mobile is a best practice that UX specialists endorse. Experts can uncover problems with this during their analysis.
However, the size that button needs to be largely depends on the context. Factors like the average age of the target audience, the circumstances in which users browse the site and how complex the site is to use all need to be taken into account. These questions can’t be answered without comprehensive user research.
While we advocate using data-driven research techniques, a human-led evaluation of a website is a useful starting point for conversion research.
Experienced experts can spot real problems, but they still need to validate these with actual user data. The biggest risk in a heuristic analysis is the confirmation bias. This is the tendency to subconsciously search for and favour information that confirms our own beliefs, without collecting sufficient data which could disprove premature theories.
- It’s quick and relatively inexpensive if you have an experienced optimisation team working for you. It can be more expensive if you need to outsource the experts.
- The method can be easily repeated after each series of changes to holistically evaluate progress.
- Experts can evaluate before the site or changes are launched.
- It can be carried out before asking real users to test your site. This means you can solve the biggest usability issues before investing in user testing. Not only does this save money, it gives users the chance to discover the subtle issues specific to their workflow.
- It’s unlikely to be valuable when used alone, as it requires other methods for validation.
- The evaluation is prone to cognitive biases.
- You need a group of experts – one person will not suffice.
Heuristic analysis vs. usability evaluation
You may find that many experts use these terms interchangeably. Although we acknowledge there is cross-over between the two techniques, we differentiate between them and use both in CRO projects.
In a heuristic analysis, specialists evaluate a website against numerous principles of different natures in order to find profit leaks. It follows a structured approach, but can be much more fluid because it’s mostly guided by someone’s expertise. It’s also prone to subjectivity.
On the other hand, a usability evaluation will focus solely on usability principles. It often consists of checklists set by authorities in usability engineering, such as Jakob Nielsen and the Scientific World Journal. There are plenty of comprehensive usability checklists you can use too.
How to do a heuristic evaluation
As with any discipline, different people have different approaches to this technique. Some of the well-known ones in CRO are MarketingExperiment’s conversion formula and WiderFunnel’s LIFT framework. At Worship we use an adapted version of ConversionXL’s heuristic framework, as detailed below.
Take each page of the website and envisage the buying stage the user is at on that particular page. You should also account for the different segments you want to optimise for, such as first-time visitors and loyal customers. Remember: you can’t optimise for everyone. It’s best to prioritise the segment that the website was built for or the one that will be most profitable.
Using the information you have about your target segment, evaluate each page against the following principles.
People fear the unknown, and they won’t feel confident buying your product if they don’t fully understand it. Your website should recreate the experience of the customer holding the product in their hand as closely as possible.
Customers should be able to quickly find answers to their questions as though they’re speaking to a friend who’s already using your product. This way, they’re more likely to buy it.
From our experience, including clear copy about what you offer is miles better than psychological tricks or persuasive copywriting.
In order to assess the clarity of the website, ask the following questions:
- Is it obvious to all users within 5 seconds of landing on your site what the site is about? Is it clear what products are available and how you differentiate yourself from the competition?
- Is it clear what page they’re on and what they can do here?
- Is the main call-to-action obvious?
- Is the copy easy enough for your target audience to understand?
- Is the copy rich in detail?
- Does the visual hierarchy of the design elements complement the hierarchy of the copy? Does it enhance the understanding process?
- Does the design help the user quickly find the information they need with attention directing techniques? For instance, are elements that talk about the same thing visually grouped?
- Is it immediately clear to the user how to achieve what they want? For example, how to add to basket and how to complete the checkout process?
- Is the font size adequate?
Throughout the journey users take on your site, they should feel like they’re in the right place and en-route to achieving their objective.
Use these factors to assess the relevancy of the site:
- Does the landing page match the site the user came from (e.g. search results or adverts) in terms of language, design and style? Will they be able to recognise it’s a continuation of their journey?
- Is the relevant information presented at the right time without them having to navigate to different pages to find it?
- Does the photography reflect your product accurately and show it being used in the context your visitors use it?
- Does the copy match the words in the user’s head when thinking of your product or their problem? If you use different language to your customers, they might not realise you’re talking about the same thing. Creating a sense of familiarity is key.
- Are you attracting relevant traffic and targeting the correct keywords? Your bounce rate will tell you a lot about this.
Getting to the bottom of what motivates your users and understanding their buying behaviour is a core component of Conversion Rate Optimisation.
During the evaluation, you can determine whether your website is designed with these motivations in mind. Ask the following questions to assess this:
- Are there clear differentiators from competition?
- Are there discounts, offers and other incentives?
- Is the website engaging and fun to use? Is it pleasant to browse?
- Do users feel they are getting good value for money if they buy from you?
- Does your copy highlight the benefits of the product (not just the features)? Do they feel your product is going to make their lives better?
- Are you using emotional targeting in your copy?
- Do you have strong social proof (e.g. rich, positive and diverse reviews)?
- Does the user immediately trust the security of your site?
The evaluation process can highlight areas of the website where users encounter friction.
Ask these questions to identify where people may become frustrated with the site:
- Could users be having doubts or uncertainties about security, privacy, trustworthiness, delivery and quality?
- Are there any questions left unanswered about your product or service (e.g. the delivery process)?
- Are there usability issues like difficult forms, lack of accessibility, poor readability, poor contrast ratios and errors on touch devices?
- Does the user need to click more than necessary to achieve their purpose?
- Is the system slow and irritating?
- Do you lack clear reviews and security seals?
- Is your contact information difficult to find or missing?
- Does the design need to be updated?
- Do users find the copy too difficult to understand?
Any element that isn’t directly helping the user achieve their desired task is a distraction. We all face constant interruptions when purchasing on the go, from people speaking to us to notifications appearing on our phones. We have short attention spans. We’re constantly in a rush. This means ecommerce websites need to minimise diversions to help people purchase quickly.
Consider these factors when assessing distractions:
- Are you presenting recommended products too early, making the customer jump from product to product?
- Do people have too many options that make it harder for them to make a decision?
- Are there any animating elements that aren’t related to the current task (like sliders and ads)?
- Are there social media attractions that don’t contribute to the purchase? Once people click into Facebook, the land of distraction, they might be gone forever…
- Is the design too diverse in colours and shapes, thus lacking a clear visual hierarchy?
- Is there too much irrelevant information?
- Does the call-to-action fail to stand out immediately because of colour and position?
- Are you interrupting the user’s journey with irrelevant pop-ups?
These are just some of the questions you can ask when approaching a heuristic review. With more practice and experience in the industry, you can uncover more problems with a website during the heuristic analysis.
Bear in mind that this is only one aspect of user research and shouldn’t be used as a substitute for any other method. Think of it as a starting point in the conversion research process that could help you spot the biggest issues on a website. And remember: always validate your findings with real user data.
25th October 2017