Why a UX audit is your secret weapon
Most people think UX audits demand endless resources and weeks of user interviews, to many clients and fast paced marketing agencies they seem like an avoidable use of budget that could be spent on ‘quicker’ wins. But a UX audit is more than just a critique on design, it’s an valuable ROI health check.
So what if there was another way, one more suited to those with limited resources?
Let us introduce you to the Lean UX Audit. Using high speed quantitative data and expert heuristics you can find 80% of your problems in 20% of the time.
Step 1: Define your KPIs
Before looking at a single screen you need to know what is important to your website and how to measure it. Think about if you’re trying to sell something, capture information or simply inform your user.
Once you know that use metrics to decipher the information you have:
- Conversions: Can they complete a task (purchase, sign up)?
- Bounce Rate: When do they jump straight off the page?
- User Engagement: Do they stay on page or scroll far down the page?
- Task Success Rate: Can they do it?
- Error Rate: How often do they fail?
Step 2: Resource allocation
Qualitative research is the gold standard but if you don’t have time to perform user interviews the next legitimate step is quantitative. Use analytics to find the fire and a heuristic review to make an educated guess at why it’s burning. You become an advocate for the user, always use data to remove bias and back up your thinking.
Tip: get more than one perspective, ask your colleagues for their opinions.
Step 3: User needs & personas
Use the data you have access to and build Personas.
- Research the ‘Need’: What problem does this product solve?
- Demographics: : Are they “Mobile Browsers”? How old are they?
The goal here is to get inside the user’s head without needing to interview them.
Step 4: Mapping the “as-is” journey
Pick a task on the website that the user appears to struggle with and go through it from the perspective of the persona you created. Start at the first touchpoint and at each one until the task is complete and write down how you ‘think’ they are feeling or performing.
Use a table with the touch points listed across the top and user metrics down the side. Some good starting metrics are: User Actions, Goals & Experiences, Feelings & Thoughts, Opportunities, Pain Points.
Step 5: Analytics and heatmaps
Now you can use the data you have at hand to validate or invalidate your journey map by crossing out assumptions and replacing them with facts.
Google Analytics: Use “Path Exploration” to see where users are actually going compared to where you assume they should be going.
Hotjar: Use heatmaps to see if users are finding the information they need to complete a journey or recordings to find out why they are rage clicking.
Step 6: Heuristic Review
A heuristic review is a walkthrough of the site based on established usability principles. It’s worth checking out Jakob Nielsen’s 10 general principles for interaction design for a full checklist but here are some I consider most essential for this exercise:
Visibility of system status: The user should always know what’s going on (e.g., loading bars).
Match with the real world: The language should be jargon free.
User control: Can the user exit or go back if they need to.
Consistency: Are touch points like buttons consistent enough so the user knows what they are?
Error prevention: Does the form prevent them from submitting the wrong information (usually an error message) before they hit submit?
Aesthetic and minimalist design: Is the design cluttered with unnecessary information?
These will highlight the why behind the analytics data.
Step 7: Pulling it all together
Now that we have some evidence backed up insight we need to present it in the right way. Instead of just writing out your assumptions use this formula:
Analytics data + Heuristic violation = Issue to solve
E.g. Hotjar is showing that 80% of users don’t click through to the sale (data), because the sale button is just text and doesn’t look like a button (Heuristics). Make the button the same as other buttons on the site.
A final note
This is a great method as a starting point for any UX audit but it isn’t gold standard. We should always try to use good primary and qualitative research to back up our findings. Speak to users and remember UX is a marathon, not a sprint: audit, fix, measure, repeat.
Pro Tip: Don’t audit everything. Follow the money and focus on these journeys – the issues fixed up here most likely fix the rest of the site.
