The Simple Way to Conduct a UX Audit

Here I'll introduce you to the 5 core parts of auditing — which just means evaluating and analyzing — an app, website or enterprise software system.

Now — you’ll see and read about lots of other ways of doing this, and that’s fine. But I don’t care about anyone else’s method; this is my way, what’s worked for me. It’s a combination of methods I’ve found to work best over the last 20+ years of my career.

I always start with the business side first, for one very critical reason: if the business doesn’t see how improving UX gets them what they already want, you have almost zero chance of convincing them to do that work.

Yes, users matter. Yes, our job is to advocate for them, ruthlessly. But we also have to live and work in reality. And reality is that too many people we all work with still don’t understand the value of UX. So the only way you’re getting a green light for the work you want to do is if they can clearly see HOW what you’re proposing gets them the outcome they are already after.

Capisce?

Here’s where things often get difficult, because far too many organizations:

1. Don’t really know what their users or customers actually want.

2. Have used faulty methods like surveys or their NPS score to ascertain the value of the product to those users.

3. Aren’t willing to allow you to talk to those users to find out what they actually need or hope to accomplish.

And it’s these situations I’m going to focus on here. We’ll assume your client or the company you work for is one of many UX immature organizations who has no research and won’t pay for any either.

What do you do? I'll show you.

The next place we look for evidence of user satisfaction or frustration is to existing analytics — specifically, we'll walk through the following metrics:

• Completion rate

• Form abandonment

• Drop-off point

• Time on page or task

• Search queries

• Error messages

• Mobile vs desktop statistics

• Unpopular screens

I'm also going to give you a few cautionary tales about relying too heavily on numbers — specifically performance metrics and Key Performance Indicators (KPIs).

Here's a simple, easy-to-follow walkthrough showing how Google Analytics' Event Tracking can be used to investigate Website form validation errors. I'll show you where to look and what to look for.

Next, we turn our attention to three states of interaction:

1. Blank State. This is what people see the very first time they launch it or when they log in.

2. Working State(s). This is what people see and interact with during the normal course of use, with data, content and controls in place.

3. Error State(s). This is what people see when something goes wrong.

When you’re conducting a UX audit, you have to look at all three, mainly because each represents an area where people may either (1) get confused and not understand how to proceed or (2) get mad and leave and use something else.

Like it or not, the expectations you set (or don’t) in the Blank State — where everything's waiting to be populated — set the bar for the quality and usefulness of the experience. It’s a first date where that person decides whether to actually go out to dinner with you or not, based on a cursory 5-second look. The stakes couldn’t be higher than in the blank state — so that’s where we begin our audit.

Now we evaluate the Working State: what happens while people are working with the product in question: viewing, interacting with and manipulating what they see on the screen — and what happens in response to those actions. We analyze the quality of the UX across 4 key principles, all of which inform the principles in the Working State videos that follow:

1. Control

2. Consistency

3. Context

4. Corroboration

Predictability means that good UX and interaction design should set accurate expectations about what will happen – before the user taps, swipes or clicks on something. When it’s easy for people to accurately predict the outcome of their actions, their confidence increases, which encourages them to continue using the product. The better the system serves their needs, the more efficiently they can use what’s in front of them.

In evaluating Progression, we lean on the principle of Progressive Disclosure, which means that everything in the UI should progress naturally, from simple to complex. So in this step, we're looking for evidence that only the necessary or requested information is displayed at any given time. Doing this makes the product easier to use.

Natural Constraints prevent users from making too many mistakes — so in this step we're looking to see whether or not the system and its interface are designed to minimize user errors. In addition, when work needs to be performed in a specific order, or where one action limits or expands the choices available in the next step, the choices available onscreen at any given time need to be limited or constrained appropriately.

In this step we're determining whether interactive elements — anything necessary to allowing a user to take action — are clearly visible (Visibility), presented in order of relevance and importance (Hierarchy) and with their availability and function being obvious to the user (Visual Clarity).

The next step in the audit is to evaluate Flexibility. Which means whether or not the app, site or system in question accommodates both novices and experts, and keeps user focus where it belongs: on the content of the interactive product and the task at hand, instead of on format, navigation or time wasted figuring out how to use the UI.

Feedback communicates the results of any interaction, making it both visible and understandable. Its job is to give people a signal that they (or the product) have succeeded or failed when performing a task. In this video I'm going to walk you through the questions you need to answer to determine how well the app, site or system you're auditing does that.

The Error State is what people see when something goes wrong — and if those error messages aren't present, clear and instructive, this can be where a large number of users jump ship, never to be seen again. In this video I'll show you what attributes to look for to determine what makes up a positive user experience: how well errors are explained, how well any system responds to those errors and how easy it is for a user to understand and correct those errors

A recap of all that's come before, and a reminder that there is always more than one way to do anything.

The UX Audit Workbook is a detailed, comprehensive, step-by-step guide that details 170-plus individual elements and attributes you need to pay attention to and analyze during your audit.

In this portion of the course, I'm going to walk you through a UX Audit, exactly as I would do it for a client. This video and the ones that follow are exactly what I deliver to clients: a video walkthrough explaining what I found, why these things are problematic for them, and what they should do to fix them.

In this video,  I'll point out a number of cognitive disconnects that hamper a user's ability to predict the outcomes of their actions — which makes them hesitant to use what's in front of them...and can cause them to abandon it altogether.

In this walkthrough I evaluate the product's ability to match user expectation, according to the Law of Similarity — which says that similar items are automatically expected to have similar behavior. In addition, you'll see how even the smallest of disconnects in matching link labels with resulting page titles can lead to user frustration and abandonment.

Here we look at the product's failure to design appropriately — from language and labeling to information architecture to visual representation — for the various skill levels and understanding and background of its users.

In this video I walk through a number of instances where the system's displayed content and interaction opportunities and workflows don't match the user's context — in other words, they don't serve or speak to the reason the user came here in the first place.

In this video, I point out various inconsistencies in multiple elements of the product, all of which severely impact a user's ability to make connections between actions/interactions, context and content relationships — as well as their ability to access and understand it.

Here I'm calling out issues with color, contrast and fonts that affect the product's scalability, readability, legibility, comprehension and the user's ability to identify functionality and understand the hierarchy of content.

This is a sample scoring summary that I'd give an actual client, based on the criteria from the previous walkthrough videos.