7 questions analysts must ask users before visualizing their data
The days of static dashboards and reports are gone. Data visualization tools make revealing insights easier than ever before… but you need to distinguish the signal from the noise.
WATCH: Unleashing your data through UX webinar
The capabilities and flexibility of data visualization tools like Power BI are truly game-changing, but that power, flexibility and opportunity can be overwhelming for users. Analysts are in a unique position to shape data visualizations to help the user find actionable insights.
That’s where Business Intelligence Engineers come in. With the goal of helping drive data-driven decision-making, it is the BIE’s job to distinguish the signal from the noise for a company. To do that well, an analyst must learn and understand what end-users really need. Reports don’t exist in a vacuum — an analyst can create reports all day based on any data, but unless those reports provide some use or meaning, they will fall flat. It doesn’t matter how great a model is, how much data is available or how pretty the design of a dashboard is. If the end-user can’t access the insights, if they can’t quickly understand the state of their business, if they can’t quickly and easily identify the story the data tells, then the report is pointless.
User-centered design is at the core of every report analysts build. To that end, these are seven questions every BIE should ask when building reports. This ensures the process is efficient and results in reports that actually answer the questions the users are asking.
#1 What actions do you want to take based on the report?
People want to look at their data and take actions. Retailers want to see which of their stores are missing sales goals, and they need to see why, so they can get those stores back on track. By understanding and identifying what they want to do with it, the analyst can craft exactly what the user needs.
#2 What type of charts do you currently get your insights from — or — how do you picture this data being visualized?
Have the user pull up what they currently use, or if they don’t have anything, have them draw a picture. Find out if they like bar charts, line charts or pie charts. Different people work best with different types of visualizations, and they may already have an idea of what they want. Knowing that upfront is beneficial because it prevents costly potential rework.
#3 How do you share this report with other people?
Whether a report is for scheduled business reviews or the user needs to view it on a mobile device, take a screenshot, circle something and text it to someone to ask what is happening, knowing the specific usage details early allows the analyst to determine the best format for the user. If, for example, the user plans to put a report into a PowerPoint slide later the analyst can set those aspect ratios to eliminate headaches. Or maybe the user always has to print out the reports — the analyst can make sure it is in a letter format, making it is easier to print and read. If an analyst knows a user is going to export something to Excel to send to their manager, they can just create the report in Excel for them connected directly to the dataset, eliminating those unnecessary steps and making it reusable every day. Asking the questions and finding the pain points allows the analyst to design specifically for any problem areas.
EXTRA TIP:
Analysts should try to get involved as much as possible. Beyond asking how a report will be shared, they should try to see it in action. Sitting in on meetings allows an analyst to see what the users need and how they interact with the data. The users don’t always know all the capabilities of Power BI or what analysts can do with their data; they’re looking to the analyst for their expertise.
#4 How much time do you spend looking at the report?
There are often two types of report viewers. There are those who spend a lot of time studying the reports, drilling down into the details and validating and understanding different business drivers. The others are usually at the executive level, and they want to quickly see how much money the company is making or how critical divisions are performing. By asking how much time a user spends looking at a report, it helps the analyst know the information they need to surface quickly and how much detail should be included for different audiences.
#5 How do you use the report?
Data visualization and dashboards are often thought to be consumed by someone sitting at their desk on a computer. But that is not always the scenario. The organization may have a team of workers on mobile devices out in the field or managers on the salesfloor who need to quickly check shift resources or inventory data. Knowing the planned usage for a report allows an analyst to suggest different options to meet the user’s needs and expectations.
#6 What is the user’s hypothesis of what the data will show?
For data visualization, the analyst needs to understand the data extremely well. The data has a story that it is trying to tell. But the analyst also needs to know what the user thinks that story is. This can help uncover data quality issues or help level set expectations.
It can be an awkward conversation to have with clients because their hypothesis may end up being wrong, but they almost always want and need to know that. These discussions can lead to a data quality issues report that provides actionable insights to fix these issues, improving all the companies reports.
#7 What does the data mean to the business?
Each number in a report is going to speak to some part of a business: expenses, volumes, number of tickets, etc. Understanding how those numbers change over time is important. Knowing how often data is ingested/refreshed — every 10 minutes, every 3 days, once a month — and how much data there is, 1 million rows per year, for example — drives the architectural design behind the reports and enables the user to see updates when they expect to see them.
The bottom line is that if an analyst doesn’t address these seven questions upfront, they can become costly pain points later. Now that you’ve seen how Blueprint builds a user-centered data visualization experience, let’s have a conversation about your business’ needs and how we can help you build your future.
