Man putting together a puzzle on a wall blocking him from a park with people
Digital Strategy

Gaby Kenyon Q&A: Bringing UX Expertise to Enterprise Applications

Posted: March 28, 2017

User experience has been generating a lot of buzz in today’s digital landscape, but it’s sometimes seen through a limited lens. Arrow's User Experience Evaluation Tool addresses an essential, yet often overlooked need among enterprises: Developing easy-to-use business-facing applications. This instrument will let you and your engineers review the UX of your enterprise apps to identify places for improvement.

The Arrow Marketing Team had the chance to sit down with Gaby Kenyon, one of our UX analysts, to discuss the evaluator tool's creation and some of her thoughts on the state of enterprise UX today.

What was the inspiration behind the UX evaluator tool?

We recognized a need for a tool that focused on enterprise applications. If you go online and search for any sort of UX evaluation, you’ll find tools that show you how to evaluate the user experience for public-facing websites. Those tools don’t take into account the complex requirements for business process automation and expectations that come along with a more focused, niche audience.

For example, it's really easy to find information on how to optimize an e-commerce website's UX, but there's not much about  improving internal-facing software.

What was the development process like? How did you turn the idea into a tool?

Ryan, our managing partner, thought of creating a UX tool specifically for enterprise applications. He thought 'Why doesn’t this tool already exist?' So, I started working with our creative director. We asked ourselves 'What do we have now? What needs to change?' We already had a tool that helped us evaluate our clients' website usability, so that's where we started. It had about 150 questions that helped determine whether a site's UX was poor, average or great. 

From there, we sat down with the UX designers and software architects here at Arrow who work on our B2B projects (I'm usually working on B2C projects) and picked their brains about what they thought the tool should include. I did a lot of research on enterprise UX best practices and how to evaluate enterprise applications in layman's terms.

Based on your research, which UX pain points are associated with enterprise software that aren't applicable to websites?

One of the biggest concerns associated with enterprise applications is that once a company has bought a solution, the end users have to use it and they have to use it every day.

Say I’m a customer on an e-commerce website. The site has poor UX - I can't find the products I want, it takes too long to check out, etcetera. Ultimately, I’ll leave and make the purchase at the site with more optimal UX. But if you have a tool that a doctor or accountant has to use as a part of his or her job, and that tool has usability issues, that professional will have to deal with those issues every day. Considering that, it's necessary to minimize the number of UX problems within an internal application.

How did that pain point affect the type of questions you included in the UX evaluator tool?

We focused a lot on user workflow. For example, we have questions that address challenges such as bulk editing. If you have to edit multiple records at once, you don't want to have to click something 70 times to address all the revisions you want to make. Overall, we tried to position the questions so they would help analysts address user efficiency.

Do you think UX often falls victim to subjective opinions? Can one judge an application based on its "look and feel" alone?

That can absolutely come into play. We tried to word questions so they aren’t purely subjective and we provide empirical evidence as to why those questions are important. It's really tough to write 'Yes/No, is this good for UX questions' because there are very few hard-and-fast rules in UX. You may have an idea that would be impractical if implemented in one project, but fantastic in another application.

We found information from multiple sources to see what fit with enterprise applications. We didn’t just sit down, spin up a bunch of questions and say 'That sounds right.' We compile research from some of the most reputable UX sources and cite them in the custom report we provide.

What mistakes do developers make when testing UX?

It depends on how the company is set up. Here at Arrow, developers don't test our UX. We have a quality assurance team that conducts all of the UX testing outside of development. After they're done assessing the applications, they go back to the developers and say 'This is what's working, this is what isn't.'

I think another big concern with testing is that the folks assigned to the project haven't documented how the application in question is supposed to work. In instances such as this, testers trying to evaluate the UX, don't have a benchmark from which to base their analysis.

Most importantly, testing without users really misses the point of UX. If all of your testing is done in-house, and you never show the application to the end-user (whether through a beta program or some early-release cycle), that's going to make your job a lot more difficult.

Alright, so you've explained how the UX evaluator came to be - let's get into the nitty gritty.

The UX evaluator tool gives considerable attention to search and documentation - Is this an area of UX testers often disregard?

With a standard website, the search function deals with pretty standard queries, but in an enterprise application, there are several very specific tasks certain people have to complete. There's also a lot of information associated with each of these tasks, so the queries are a lot more complicated.

Add this to the fact that when a business hires a new worker, that employee needs to quickly familiarize him or herself with the application. Good search functions and documentation can go a long way in helping that worker figure out how to use the software with as little instruction as possible.

How important is metaphor composition when developing a system's UX?

Not every system needs or uses metaphors. In the broadest sense, if you have "folders" in your application, you're using a metaphor that helps users understand that they can organize files according to topic, client or another classification. Problems arise when you use metaphors that don't correlate with the user’s mental model of a particular object or function.

What is Fitts's Law, and how does it factor into user efficiency?

Fitts's Law states that "the time required to rapidly move to a target area is a function of the ratio between the distance of the target and the size of the target." Essentially, it says that it's going to be easier for a user to click on something if that something is large and close to what they're trying to control.

Let's say you have a graph that represented the number of users who visited your website overtime. You can change that timeline to a week or a month by clicking different options. You expect that control to be close to that graph. If your graph is on the left side of the screen, and the options are all the way to the right, it's going to take the user a lot longer to find that timeline function.

In your experience, do UX researchers often ignore the importance of writing concise error messaging?

The problem isn't so much concise error messaging, but writing specific error messaging. Let's say you're filling out a form to register for a service. You hit submit and get an error message saying "All fields are required." This implies you forgot to fill out certain fields, but doesn't tell you which fields you missed.

How do you believe the evaluator tool will evolve in the future?

I think that we would like to be able to use the tool to assist our B2C clients as well as our B2B clients. Right now we use one of those pre-existing tools for evaluating B2C websites with our clients. But I think using an asset we developed in-house that's really specific to the problems we encounter could be even more valuable. We all know how fast UX changes in the B2C space. If we could adjust the UX tool to possess this level of flexibility, that would be amazing.

I also think there's an opportunity to use this as a UX tool that enhances specific business processes, whether they be in health care, finance, or any other industry. It'd be interesting to develop a tool for each vertical, or perhaps one associated with each application. If you walk into your doctor's office and take a look at the tools they use - they're right out of the '90s. They're using really old systems, and they don't have much of a choice. That represents a huge opportunity to improve the experience of people using enterprise applications in their job every day.