UX audit. Self-service guide & few rad resources to make things rolling
- what UX audit is,
- when it needs to be done,
- what it gives and what sources,
- and methods of auditing to use.
Keep in mind, that GRIN tech’s digital marketing & development blog is focused on delivering content from Entrepreneur & Business perspective, so this article is tailored accordingly.
What is UX audit
I think for digital industry it is quite common perception that UX is about improving easiness of website use (i.e. helping users to perform some action) via design changes. It is more beneficial to assume a wider perspective – e.g. intrusive (or broken) email automation is also a User Experience in its core and therefore the whole set of user interactions with brand via Internet can be subject to audit and re-evaluation.
When to perform UX audit
The basis of the basics is setting goals. Take a look at the site and answer the questions:
- What is it for? What is the value for the user?
- What do you want to achieve by auditing? How will you evaluate the result?
- What resources to use to introduce improvements?
Choose the page for UX-audit: look for the primary “entry” pages on your website, select one of the most priority to you. Analyse its key metrics (number of calls or redirects, registrations, time spent on the website, rejects) – they will be needed to find out the reason why the page is in decline. Then, select your source and method.
Proper UX should address some specific issues. This is an opportunity to find out what users are experiencing (mainly on website) and why they are going away. The resolution of these issues will have an impact on their loyalty and their ability to go from potential to active customers. You are going to also strengthen your market position over competitors. The more convenient your website is, the more likely the buyer will use it as intended.
Decide whether you need auditing based on metrics analysis. It’s time for UX audit if:
- users visit the website, but do not perform targeted actions;
- users leave in the process of purchasing the product;
- users spend little time on the website and do not go beyond the first page;
- interface is overcomplicated: useless sections and categories, unclear functionality.
If the website has technical problems – fix them. If there’s no traffic – find out why. But do not bet on UX audit. In these cases, it is useless and even harmful. You will only spend time on improvements that are not the root cause of problems and still will not get the result.
UX audit tools
Here is our usability testing tools list for optimising user experience:
- Google Analytics. Individual analysis is important for every company. If you are not confident with a tool I recommend checking out this article.
- A/B testing. Most popular platform is probably Optimizely. Keep in mind that A\B testing endeavour is worth it only if your pages receive good amount of traffic, because otherwise you might get statistically insignificant results or improvement will be worthless from ROI perspective.
- Surveys. Standard implementation is just a pop up form on your website asking questions like “did you find what you were looking for?”. Can be done with Qualaroo, WordPress plugins and what not. If your project has a sign up process I’d also include one or two short questions there e.g. “what is the primary feature you signed ip for?” etc
- One of the less widely know approach is to run surveys via Amozon’s Mechanical Turk.
- Heat & scroll maps known as a click-based user experience tools. Popular services are Crazyegg and Clickheat but I’d rather stick to Yandex.Metrika coz its free solution of enterprise quality.
UX audit methods
UX-interface Audit: “Checklist” method
Probably you have already used checklists and this method is nothing new to you. I like this one even though it lacks interactivity, but do search for a checklist fitting your niche. For example we have an ecommerce checklist here (253 items).
- a lot of items: a good checklist takes into account all the interactions on the website, does not let you forget about the details;
- no skill required – you make corrections based on ready-made recommendations.
- it is difficult for a specialist to determine the quality of the checklist;
- items in some checklists raise questions: they are categorical and do not take into account the diversity of business situations;
- There is no explanation on reasoning behind a statement of the checklist.
I would recommend using checklists after the audit is done to test the performance. Do not choose them as a stand-alone method.
Heuristic evaluation of UX interface
Heuristic evaluation of the interface was developed in 1990 by Jakob Nielsen (actually, Rolf Molich was also involved). It is thought that the method should determine the UX-interface problem, based on the human factor. There are 10 heuristics in total. There are no scientific proof behind them and UX-specialists still argue about their use.
Audit the interface for each of the heuristics and correct what cannot withstand the criticism of the method. The method is like the “Checklist”, but heuristics are less definite.
- a free and fast way to evaluate the resource.
- a popular method. Ask any UX-designer and they will confirm – the method is “alive and kicking”.
- heuristics are not proven scientifically: in fact, you conduct an audit on personal recommendations of an expert.
- you will not find errors outside heuristics
Find 3-5 experts in UX design and ask to audit the site. In a two-day period, experts will find 95% of bad spots and file you a report.
- experts will make the UX audit faster than you: they know the main errors and notice what escapes the sight when auditing with the focus group;
- most likely experts are knowledgeable in SEO, programming and God knows what: you will get a comprehensive analysis and ready-made solutions for all problems.
- Experts are subjective. Their point of view is likely to differ from the users point of view;
- There is a possibility of an error in determining the root cause of problems in the interface;
- Experts may be inexperienced.
Expert opinion should be used in conjunction with methods that involve research of user behaviour.
“Locate Google’s top 5 sites of your competitors. See how and what features they implemented. Copy. Do not come up with the impossible and the unique: place only the required data on the website. You should do not do as you want, but as the user wants. For a layman it is better to apply the focus group method: ask 10 acquaintances to draw what, in their opinion, your site should look like ( with content). And then match the score with the websites of your competitors”, – Eugene Guschin, UI designer.
Card Sorting Method
This method allows you to evaluate the usability of navigation, the categories and the location of the features. It is simple, but at the same time effective: it will be necessary to involve users of the site to analyse their behaviour in real time. From 7 to 10 people will be enough for the test group.
Take screenshots of site elements in advance: button, menu, navigation links, captions, and so on. Print them, cut them out in the form of cards and issue them to the members of a test group. Now invite users to sort the screenshots according to personal ideas about the perfect structure of your website – at this stage you will already see interface errors in terms of usability.
Go to the second stage of card sorting: use all the options of the method or the one that matches the goal. There are 3 options: open, closed and reverse.
When sorting is open, combine the cards into categories (for example, “headers”, “arrows”, “forms”), names the categories to the users. In turn, show the objects to the main group and ask for a name of the category to which they belong. The check indicates whether the set of categories is valid for the layout of items on the page.
When you conduct a “closed” sorting: the audit participants come up with names for groups of objects themselves. This practice makes it possible to see the connection between the elements of the site through the view of the users.
When you “reverse” sort, you check the structure of the hierarchy. The person needs to find the right element on the card by scanning the page with a look from the top category to the bottom. This method will detect a problem of navigation or structure of the website.
Record test group audits on video: careful analysis will help in preparing UX interface improvements.
Component Expertise Method
The method gives a complete view of the current situation: after the audit, critical errors can be corrected and new elements can be introduced to improve user interaction. “Exposure of components” is time-sensitive: if you’re not going to be able to take a long time for research, it’s better to delegate the task to specialists.
Before proceeding with the audit, form a general opinion about the target audience. If you have a lot of information, create an average profile of different groups of users. Then, specify the context of the use of the website and 3-5 behavioural scenarios for each type of user.
Ask questions for each scenario:
- How to make the work of a user with the site faster?
- How do I remove or reduce the number of errors?
- How to increase their satisfaction with interaction?
- How to increase the speed of learning how to interact with the resource?
Of course, there may be more questions.
Then, enlist the focus group to the audit: get each person to go through the ready-made scenarios. Record the results, conduct an in-depth interview. State the problems of the interface, record the possible solution. In parallel, setup Google Analytics on the website and setup goals. Implement changes gradually and, as you do, analyse the achievement of goals.
To make it easier to follow the sequence of actions download the template for auditing.
Naked self promo
A ritual exchange of money for solutions one of my favourite things to do. So, since among other things GRIN tech is UX & UX agency why not outsource this task to us completely?
To sum it up on UX audit
Personally, I believe that checklists & experts can get 80% of the job done at 20% cost. The last mile that is not covered is worth pursuing in case of huge traffic \ sales volume, when even a 0.1% increase in conversion rate will result in significant cash.