Every organization has a unique reason for existing, or else they don’t exist for long — they make an interesting product, deliver stuff faster, or provide better service. The ones that continue to exist also have a plan for how they will provide that value. This is called strategy.
A well-articulated strategy helps people in an organization know what to do and how to do it in a way that complements the work around them. However, a strategy can’t just be thought out, it must be shown and reinforced with the people responsible for doing the work. This is why visualization and strategy are perfect companions.
Successfully executing a strategy requires getting many people, all with different roles, to agree on a common goal and cooperate to bring it about. Visualization, usually in the form of an interactive dashboard, fosters agreement and understanding by distilling complex ideas down to the essentials. It also guides groups of people toward the right activities, distributing the message widely and presenting customized views to each user based on their specific role. A great strategy needs a dashboard to guide people towards the right activities. A great dashboard needs a strategy to give it a useful structure.
Unfortunately, many dashboards fall short of this ideal. Sometimes the problem is fixable, such as poor visual design or slow performance. But often it is caused by an overall approach that is harder to fix without starting over. This happens when you start creating visuals before you get clarity on the strategy.
While strategy and visualization work hand in hand, the sequence of the two matters. In a dashboard-first approach, you cobble together whatever metrics happen to be available into a few visuals and then search aimlessly for a problem to solve. Without the guidance of a predetermined strategy, its usefulness is limited, and if it does inspire users to act, the actions have little connection to a larger goal.
A better way is to take a strategy-first approach where you first determine what the organization wants to do, how they plan on doing it, and why they should care about it. Then you gather data and design visuals to that end. Or to put it another way,
Before you visualize, know the what’s, how’s and why’s.
I recently worked with a client who was responsible for tracking how well new product introductions were resonating with retailers. The process was historically haphazard, and metrics had varying levels of detail and accuracy depending on the customer and product, which made production planning a challenge. His job was to bring structure and clarity to the process with better data collection and visualization of the results.
As we began formulating the dashboard requirements, we quickly realized that there was little consensus among the various stakeholders on critical things like what the goal was, how to measure it, and who was responsible. At that point, we should have stepped back and figured those out. However, since timelines and budgets were tight, we kept moving forward.
With no meaningful message to communicate, the dashboard became nothing more than a conglomeration of miscellaneous metrics and hierarchies. The predictable result was an unsatisfactory experience for users and a rarely used dashboard.
If you’ve been doing dataviz for any length of time then you’ve probably found yourself stuck in a similarly frustrating, dashboard-first scenario. Here are the telltale signs…
- Endless design iterations that never seem to get closer to what users want
- You have to use 6pt font just to fit all the metrics into a single view
- Users keep asking for slightly different versions to fit their “unique” needs
- You get the always popular question, “Can I download this into Excel?”
- And, of course, the most obvious sign: no one uses it
This is by no means an exhaustive list, but if you see any of them, it’s likely time to review the purpose of the dashboard and reassert a strategy-first approach.
The Strategy-First Approach
The strategy-first approach requires more effort upfront, but it significantly reduces development time, and more importantly, produces dashboards users want to use. Instead of starting with metrics and dashboards, it starts with understanding the goals of the organization and its plan for how to achieve them. Then it determines which metrics help users take the right actions. Only after these steps are complete does the dashboard work begin. It looks like this…
Now comes the fun part. It’s time to put pixel to screen. If you’ve made thorough work of the previous steps, the dashboard practically designs itself. Ok, maybe there’s a bit more work left, but whereas the previous process led to constantly getting stuck, struggling through trial and error to find a solution, and never feeling certain, this process makes the flow crystal clear. The result is less time getting stuck and more time enhancing things like aesthetics and user experience.
Structure a view to mimic the decision-making process of users. Before a user decides to act, they generally have a progressive set of questions to answer. Understand that progression and organize the visual elements similarly. Use techniques like progressive reveal and pagination to show only the elements needed to answer one question at a time. A well-structured flow not only demystifies the decision-making process for each user but also aligns a diverse set of users around a single strategic vision.
A lot has to go right for a dashboard to work. Among other things, skills in a particular tool or language should be expertly used, good design principles need to be applied, and chart selection must be appropriate. Most projects can succeed with less than perfect execution in a few of these areas. However, no project can succeed without absolute clarity on the strategy it is designed to support.
As tempting as it is to jump in and immediately start exploring data and designing cool charts, don’t shortchange the painstaking process of first learning the objectives and strategies of the organization. Dashboards that get this right will quickly amass engaged and loyal users.
So, the next time you visualize, first figure out the what’s, how’s, and why’s.