“The gateway to achieving business value [with BI] is user adoption. You cannot circumvent that.”
- Greg Brown, Blue Margin
Does your data team understand the design techniques that will ensure adoption and improve company value? Without the right approach, your investment in BI can produce anemic results.
In this episode of Expert Insights Series, hosted by BI Consultant Greg Brown, we feature Brian T. O’Neill of the company Designing for Analytics. Brian shares his deep experience in how to create value from data projects and the impact of human-centered design on business outcomes.
“Business is people. You can’t get to business value if you don’t first go through user adoption. Business value will follow if you design something indispensable. If not, you’re a cost center. You’re just spending money.” - Brian T. O’Neill, Designing for Analytics
Designing for Analytics
Brian T. O'Neill is founder and principal of Designing for Analytics, a consultancy that helps data leaders increase adoption of machine learning and other analytics solutions through human-centered design. Over 25 years, Brian has helped companies like Dell EMC, TripAdvisor, Fidelity, NetApp, Roche, and AbbVie improve adoption and impact through UX. He also hosts the popular five-star podcast, Experiencing Data, where he interviews leading voices at the intersection of design, ML/AI, analytics, and data product management.
Watch the full interview, listen to the podcast, and/or read the interview highlights below.
- How to approach “human-centered” and outcome-focused design.
- How to “get out of the way” for great UX.
- How to design solutions that “fit” the business user.
Human Centered Design: Focusing on the User’s Needs and Outcome
“Nobody wants your dashboard. They want something downstream from that. They want the outcome, not the output.” – Brian T. O’Neill
Output vs. Outcome
Historically, BI projects have focused more on the reports to be produced than the business improvements to be realized. As a result, surprisingly few companies can “consistently embed analysis, data, and evidence-based reasoning into their decision-making process” (Deloitte, 2023).
This disconnect stems from a myopic focus on outputs which misses the greater purpose of BI – equipping end-users in their pursuit of better outcomes.
As Brian puts it, “We’re in the business of aiding decision-making.”
Design for Outcomes
Outcome-focused design targets the end-user’s actions in support of a broader business objective (Head, 2023). So, defining those objectives for a given user group is the right place to start. And teasing out the essence of those objectives is best done through questions. For example, if the desired outcome is increased revenue, ask your SME or stakeholder: What is our current revenue? By how much do we want to increase revenue, and by when? What leading indicators most reliably correlate to increased revenue? As Brian says, “You do not want to find out how to keep score at the end of the game. It’s much more fun at the beginning.”
Focus on User Needs
Once you’ve defined the desired outcome, the timeframe, and the leading indicators, start designing your dashboard with the end-user in the leading role.
Do this by asking questions like:
- Where does the individual most influence this outcome, and how is their contribution measured?
- What’s their process for getting to points of impact, and what parts of that process create the most value?
- What information do they need to understand whether they’re winning, why they are or aren’t winning, and how they can turn the tide? In other words, what narrative gives them clear oversight, then shows them what decisions, prioritization, and action steps will most improve outcomes?
User Experience (UX) Design: Getting Out of the Way
“When was the last time you thought about the interface on your phone’s calendar application? You’re not thinking about how the calendar interface works. It just works. That’s what we’re going for with user experience. You shouldn’t be noticing the design.” –Brian T. O’Neill
It’s a common mistake to confuse UX with “making things look nice.” The goal of UX is to minimize the time and frustration a user spends getting to their goal. As Brian says, “Good data product design is about getting out of the way.” When the tool is seamless to operate, it becomes “invisible” and is much more likely to be adopted. Following the user’s natural process of awareness, analysis, and action helps minimize UX friction.
Subtraction is Just as Important as Addition
Additionally, for teams who want to improve at “getting out of the way”, Brian suggests that less is often more. Effective UX designers recognize that every element added to a screen increases cognitive load and should be chosen with care.
When it comes to designing dashboards, if any element doesn’t evoke a question, drive an action, or inform a decision (if it’s “just informational”), get rid of it. You have to be mercenary in this practice or you’ll quickly cross the threshold of diminishing returns, where “more” crowds out “results.”
In her book Storytelling with Data, data visualization expert Cole Nussbaumer Knaflic teaches the science of UX, based on Gestalt’s principles of visual perception. She writes, “Clutter is your enemy” (Nussbaumer Knaflic, 2015, p. 36) and advocates for ruthless elimination of unnecessary elements.
Adoption Design: Using Exploratory Qualitative Analysis
“Business is people. You can’t get to business value if you don’t first go through user-adoption. Business value will follow if you design something indispensable. If not, you’re a cost center. You’ve just spent money.” – Brian T. O’Neill
BI can’t and won’t create business value unless it shifts how your people view their area of the business, and their influence over it. Adoption is the key, and if you’re looking to improve the adoption of your BI tools, you need to talk to the people using it.
The Value of One-on-One Qualitative Analysis
Brian’s recommended method for discovering how end users make decisions (and subsequently improving the adoption and impact of your BI) is the one-on-one “qualitative analysis conversation.” In a nutshell, data teams need to get good at asking questions and even better at listening.
By gaining fluency in the language of the business, data teams are better positioned to translate business objectives into actionable data insights.
At Blue Margin, we’ve found that starting with wireframes (i.e., storyboards) before engineering a dashboard is the most expedient way to understand what users need and will adopt. Sketch out their data journey, and walk through each element to determine which ones most improve users’ questions, decisions, and actions. Ask questions to make sure that their questions are answered in context, that the proposed dashboard follows their natural process, and that every element adds significant value to their process.
When end-users are involved from the outset, when they’ve wrestled with their process and the insights needed to be effective in their work, they are much more likely to adopt the final product.
Connect with Brian and Designing for Analytics
If you’d like to speak with Greg about your BI strategy, reach him at email@example.com. He’s happy to help. No strings.
About Blue Margin
Blue Margin helps PE and mid-market companies quickly convert data into automated dashboards, the most efficient way to create company-wide accountability to the growth plan. We call it The Dashboard Effect, the title of our book and podcast. Our mission is to accelerate your value creation plan.
More Expert Insights Series Interviews and Resources
- BI Projects that Drive Financial Transformation, Insights from a PMO Leader
- Trace3’s VP of IT: Accelerate Value Creation with a Company OKR Scorecard
- One Rock Capital: Top Line Growth Strategies
ResourcesDeloitte. (2023). Insight-driven organization | Deloitte Insights
Head, Kandice. (2023). Output vs. Outcome with OKRs. What Matters. https://www.whatmatters.com/faqs/outputs-vs-outcome-okr
Nussbaumer Knaflic, Cole. (2015). Storytelling with Data. Wiley.
O’Neill, Brian T. (2023). The Experiencing Data Podcast. Experiencing Data (Podcast) | Designing for Analytics (Brian T. O'Neill)
O’Neill, Brian T. (2023). Designing for Analytics (consultancy). Designing for Analytics - Design / UX / UI for AI, ML, Data Products
Spool, Jared. (2023). "Design is the rendering of intent." https://www.linkedin.com/in/jmspool/