A focus on analytics end-results creates a divide between IT teams and business users, particularly if the platform is intended to be self-service. Here’s why the road to analytics success is paved with good design
Most technology-heavy businesses, regardless of their industry, have a single goal in mind when rolling out new big data analytics projects: Get the job done as quickly as possible. Sometimes, there’s little attention paid to how the system works, but only that it does. And as many businesses are finding out, simply creating an analytics system doesn’t mean that users will find value in it, if they even work with it at all.
This emphasis on the end-result often creates a divide between the IT teams that implement an analytics solution and the business users who will be using it, particularly if the platform is intended to be self-service to any significant degree.
IT teams have one vision for what the analytics success means, how it should look, and how it will function. However, business users have entirely different opinions. The trouble is finding middle ground and meeting those desires and needs halfway.
Analytics design is the answer
Design can act as the language, per se, that connects IT and business users. Jeff Hendrickson, the director of UX (user experience) and visual strategy at Information Builders argued in a recent webinar that improvements to the design and usability of any analytics application is dependent upon getting IT and business users together in the same room. In his experience, the best ideas are discovered when these teams solve problems together.
[ Related: Lack of Real-Time Data Slows Down Decision-Making ]
From the IT perspective, Hendrickson says that by emphasizing design, “We’re showing the users that we really care about them. We’re putting them at ease by aligning with their challenges, and by doing our best to provide a solution that fits their needs.”
The goal is to use design to tie all the pieces of content together to support quick decision-making. This is supported by a school of thought known as “design thinking,” which, according to IDEO, “utilizes elements from the designer’s toolkit like empathy and experimentation to arrive at innovative solutions.”
It’s a human-focused approach to innovation, and one that focuses on the experience —particularly the emotional experience — of the user over other factors when deciding how something should look and function. It’s less about pretty bells and whistles, and more about intelligent organization of content and control.
With these ideas in mind, there are a few ways that companies looking to improve their analytics implementations by empowering them with better design:
1. Understand the necessary content
At the beginning of an analytics implementation, IT teams and business users need to synchronize about what type of content, and thus results, are actually desired. Hendricks says, “We can’t have a true design strategy unless we understand the content we need to show.” And without coming to an agreement on that content, the end-result is never going to align with the user’s needs.
One easy way to do that is to begin by filling out the following sentence: “The user needs a way to _____________ so that _____________.” By focusing on the main pain points and desires, it’s easy for disparate teams to sync up on what content they’re after.
2. Focus on simplicity wherever possible
“We’re always designing to give the user a starting point that’s clear, that’s concise, and that’s understandable,” Hendrickson says. If you throw all the information at the business user all at once — particularly information they’re not going to fully understand — they’re going to get overwhelmed. This reduces their capacity to make meaningful insights in a reasonable amount of time. According to Hendrickson, the goal is “analytics apps that provide any decision-maker with fast, easy access to actionable information without having to understand the complexity of the underlying data.”
3. Take advantage of AI where possible for analytics success
UX designers are familiar with the idea of A/B testing, wherein designers make iterative changes to the design of a landing page, for example, to test different colors on a “Buy now” affect customer response. AI is remarkably good at creating countless variations on a single design, and thus could be a valuable tool in pushing the fundamentals of A/B testing to new heights.
4. Tolerate failure
Design thinking doesn’t encourage failure, but it doesn’t discourage it either. In fact, because this methodology focuses so much on prototyping and quickly-developed iterations, it should be expected. But if failure comes often, and failure is recognized and converted into useful insights about how the platform can be improved moving forward, it can be used to benefit business users, not hinder them.
5. Think about results, not products, for analytics success
Too many companies are focused on developing the coolest new analytical model or tool —particularly one that they can show off to customers as proof of concept for their innovative spirit. Instead, design thinking teaches us to prioritize not what’s cool, but rather what will have an impact on the organization.
[ Related: Why You May Want a Career in Data Science ]
If a product isn’t getting the necessary results, despite its complexity and impressiveness, it’s more a hindrance than a meaningful contribution to the entire point of using analytics: to empower the right people with information that makes their decisions better than ever.
This is only the beginning — design thinking itself is an entire school of thought that people dedicate their careers to understanding and implementing. How design thinking intersects with analytics success is a conversation that is just beginning, which means there’s no better time to get involved.