Why and who needs dashboards?
People and companies love numbers - digitized data helps keep things under control, analyze, and make plans. It is especially convenient when someone has taken care of you and collected a lot of useful numbers in one place. It is even nicer when they are supplemented with graphs that clearly show the state of the indicators.
In this form, the figures gain power. Now it is no longer just a lot of disparate information but data that is easy to operate and make decisions based on.
Such interfaces are called dashboards or analytical dashboards. They contain reports with numbers, graphs, charts, tables, and lists. The tool is so flexible and relevant that it is used in applications, services, and programs for any sphere.
Companies need dashboards to analyze business indicators, while the government needs them to control the work. In B2C services, dashboards help solve a user's private tasks, such as tracking progress in an English learning application.
- Accumulated information in one place. Dashboards collect key summary metrics. This relieves the user from examining multiple data sources: tables, reports, lists, and registers.
- Automated analytics. The interface reflects the information ready for analysis. Necessary calculations are performed and supplemented with visuals. Much of the analytical work has already been done for the user.
- Speeded up decision-making. Raw documentation can take hours to analyze, while the dashboard has everything you need. Your key data is always at your fingertips and can be viewed in less than 10 minutes.
- Reflection on the situation and highlighted problem areas. The dashboard, like the dashboard in the car, shows the key metrics that keep the driver in control. In the same way, the user gets the big picture, and processes become clear and transparent. The dashboard clearly shows which segment the indicators have dropped and what you should pay attention to.
A competently compiled dashboard saves time, reflects the real situation in the business and individual processes, and helps to make decisions.
What do users think when using dashboards?
An important point in developing a dashboard is the scale of the user's vision. Depending on the tasks, he or she may be interested in more general or private indicators.
This is best seen by the example of the hierarchy in government agencies. For example, the dashboards for the pre-trial appeals system should have screens for three levels of civil servants:
- Deputy Prime Minister. He or she assesses the effectiveness of the system statewide.
- Minister. He or she evaluates the system's performance in terms of a single ministry.
- Inspector. He or she works at the regional level, including the specific complaints of citizens.
Each role requires dashboards with different data:
- The Deputy Prime Minister's dashboard should reflect the system's performance across the country: the total number of complaints, violation of time limits, the average time for decision-making for all subjects, etc.
- The minister's dashboard reflects the same indicators but in the context of the ministry.
- The inspector should have a dashboard with a task scheduler and private complaints from region residents, as he or she works directly with them.
Small details are not interesting to the minister, but the inspector, who works with the region, needs them. Even 10 unclosed complaints will be a reason for him to conduct an inspection. And the minister has indicators in the thousands, and they will not play a significant role in making decisions at this level.
Another great example of developing dashboards is an ERP system for a feed plant.
Suppose the owner of the feed plant wants to have the summary indicators of production, purchasing, and sales. He or she wants clear, transparent reports to make the business manageable. Such a user needs data in big strokes - it is important for him to see the big picture to make management decisions.
At the same time, the ERP system has functionality for a technologist who creates formulae for kneading fodder. He or she does not need financial reports for the work as he or she operates with the volume of ingredients, their ratio, and planned or actual production. This is more narrow and specialized data.
So, when developing dashboards, you should consider the user's needs. The same information has different relevance. For some, it will be an action signal. For others, it will be a useless indicator that cannot be used in analytics.
What should be on the dashboard?
Before you make a dashboard, you need to understand what data will be reflected on it.
In an ideal world, the customer knows what numbers employees need. And he or she is also ready to group them by priority and importance. Then the development team can skip the analytics phase and start developing screens immediately.
However, more often than not, this is what happens:
- The company doesn't have a ready-made database of metrics. In this case, a list of data displayed on the dashboard must be compiled from scratch.
- The company has an approximate vision for filling the dashboard. The initial data set needs to be further developed toward expansion and reduction.
- There is a massive amount of data available. Here it is crucial to select really useful indicators so the user does not get drowned in the flow of information.
In all cases, you need analytics - research to help define the user data pool. Below, we will list some tools that are effective for this purpose.
In-depth interviews (Customer Development) help to find out directly from the user what information is useful in his or her work and the most critical indicators. CustDev allows learning:
- What metrics do they use to evaluate the project
- What they pay attention to in their work
- What they look for in their daily routine
- Which metrics are most important to them
- What would they like to see on their dashboard
- What they lack and what they are unhappy with in the current dashboard
A real in-depth interview is when you discover how users live in the system and what they need certain indicators for. If you've done a survey but don't know the employee's daily routine, the research is unlikely to help design.
So, you should talk to those who will be using the dashboard. First-hand information will help identify the data relevant to the dashboard and cut out the unnecessary.
Another way to determine how users make decisions is to study their workflow. This method is relevant for creating dashboards for working systems that people already use.
For example, you can ask employees to spend a typical working day in the system and watch them perform routine tasks. Before your eyes, people will use the system and explain why certain actions are performed.
So we get a closer look at employees' tasks, priorities, and milestones. Ask the user to spend a typical working day and find out what is happening in their routine.
Total immersion into the user's processes will help you understand what categories the user thinks about and what data he or she operates with most often.
Exploring the decision-making process
You should engage knowledgeable consultants. Specialists on the client side are more familiar with the end users, their tasks, and the specifics of decision-making at different levels. Teamwork helps the development team create a tool based not on fantasies of how it could be but on how it actually is.
How many screens for dashboards to do?
The data pool for the dashboard is formed, and now you can apply that information in the visual context. To make it easier for the designer to shape the logic of the interface, you need to categorize the indicators into groups.
The basic principle of grouping - indicators should describe the same phenomenon or subject. For example, it can be performance indicators, indicators of a separate direction, or indicators of a division or employee.
The grouping of indicators depends on the tasks and processes of the individual user.
The resulting groups are easy to operate in the visual. If the data in a block is small, they can be placed on a separate plate on the dashboard, and if there are a lot of them - they should be placed on a separate screen.
The ideal structure of a multi-level dashboard is as follows:
- On the main screen, you place the bars with the main indicators of the data groups. They should reflect the whole picture. It is enough for the user to look at them to understand whether everything is in order in this direction.
- Additional screens reveal general information in more detail. Here the user can go deeper into the indicators and examine the picture in detail.
The principle "the more, the better" does not work in dashboards. An abundance of data looks impressive when submitting a project but is not useful for a person who wants to solve his or her problem quickly.
A dashboard should present as much data as the user needs, and not a digit more.
Multi-level dashboards are only relevant to those professionals who analyze large amounts of data. Many users will be satisfied with a single screen with a few sections.
Watch the balance of the data, and don't let the user get overwhelmed with unnecessary information. Otherwise, they will study dashboards, which should make life easier, not harder.
Rules for placing data in the interface of the dashboard
You have the data and structure - you can place the indicators in the interface. We have formed a few rules that help users quickly read the data and avoid getting lost among the numbers and graphs.
Collect data in sections
It's convenient when indicators of the same category or meaning are placed in a single section. This visually separates data groups, so the user is not confused and understands which query is answered by the information in the block.
A chart, structure, and total number can all be placed in the same section if they all reveal details of the same phenomenon.
Highlight the meaningful major
Anything related to general and key indicators should be highlighted in larger sizes. The size of fonts, graphs, and bars emphasizes and shifts the user's attention to the data that needs to be examined first.
Without these accents, there's nothing to catch the eye, and the user will spend more time looking through the entire dashboard.
Leave some air
A screen where every pixel is filled with information is distracting and increases the time it takes to find the right metrics.
Leave enough space between interface elements to keep the data from merging into a single canvas. This makes the blocks easily distinguishable from one another.
To leave more useful space for information, you need to reduce or collapse elements of the main interface as much as possible: sidebars, navigation, drop-down lists, filtering, and buttons.
Communicate with the user in color
Dashboards should help with analysis, and a color is a great tool for interacting with the user.
People have formed color patterns. For example, green is automatically perceived as something with a positive or affirmative message: ok/confirm/system works fine. Red signals that something has gone wrong.
Use bright contrasting colors
Additionally, colors help separate types of information. It's better to use bright hues in the design - this highlights the numbers and graphs in the interface. In this way, the user quickly finds the data in the workspace.
For example, if you use colors of the same gamut in this diagram, the structure of the indicator becomes not so distinguishable.
Working with color must be filigree. It's easy to get to garnish combinations that will ruin the dashboard's appearance.
How do graphic visuals work in dashboards?
Graphs, charts, diagrams, and maps are ways to graphically interpret an indicator. They are the dashboard's primary tool and help analyze data faster. By examining the screen, the user visually reads information about the situation in the business or division.
For example, the graph immediately shows the dynamics of the indicator. If you present the same data in a table or a list, you must look at the figures for each period to assess the trend.
Just try to assess in your mind the dynamics of a series of numbers: 23, 46, 15, 48, 32, 14, 18, 56, 118, 16, 20, 25. Not the easiest task for the average person.
The structure diagrams show the predominant shares. Usually, there are no absolute values in them, but visually the user understands whether this structure is adequate for the processes. If something is wrong, he or she can go deeper into the private indicators. If everything is OK, he or she can move on to the next data.
The user who operates with information by territory can use maps. This is relevant for network managers or high-ranking officials.
Where to get charts and graphs for dashboards?
There are two main approaches in the development of data visualization.
Libraries and plugins
This approach embeds pre-made graphs into the interface, refining the design and colors.
Libraries and plugin solutions are designed for simple indicators such as standard structure and trends, so the method is suitable for visualizing simple data. List of tested libraries that MaybeWorkd developers have used in projects:
- Apache ECharts
You can also use applications or search the Figma community for already rendered graphics, which will be relevant to tight deadlines. But, as a rule, Figma functional plugins and fill-ins are not known for their ingenuity and interesting solutions.
The pros of off-the-shelf solutions are saving the designer's time, the ability to reflect the most common indicators, and ease of implementation at the development stage.
The cons of off-the-shelf solutions are limitations in the variability of data and visualization if the designer plans to use a more creative approach to design.
Creating charts and graphs from scratch
When you need to visualize a specific set of data, and the capabilities of ready-made sources are not enough, it's worth developing the visuals yourself.
Creating your visualization and more complex options will increase development time since not all of the designer's ideas can be implemented in practice. Therefore, we recommend filling the dashboards while consulting with technical experts.
The pros of own charts and graphs development are no limitations on data types and visualization complexity.
The cons of developing charts and graphs are more labor-intensiveness and possible technical difficulties.
How to improve the dashboard?
Graphs save time when they are easy to read. If it takes a long time to recognize the logic of a visual, it goes to waste. Your job is simplifying the work, not breaking the user's head.
Let's move on to the nuances that will help finalize the dashboard in style and functionality:
- Animation. Animation makes the interface more lively and attractive. The best option is to add it only when the page loads so that the movement does not distract the user during use.
- Dark theme. Developing a dark theme is relevant for dashboards that will be broadcast at meetings. The dark theme looks better on large screens. Choosing a color mode will add an element of customization and expand the user experience, making the dashboard even more convenient for those who are more comfortable working with a dark theme.
- Tips. You can create a recommendation section if your organization has clearly regulated processes. It will display an action plan for an employee in case of deterioration of one or another indicator.
- Pop-up figures. It is convenient when pointing at the diagram you see explanatory information about the indicator that you do not have to search for in other parts of the dashboard.
In dashboards, you should choose the simpler and more obvious types of diagrams that most people have encountered in school, college, or work. The brain quickly considers the familiar logic, and the person doesn't have to spend time learning to understand the information on a new chart.
Creativity will be appropriate in the charts' design, color scheme, and elaboration of shadows and other elements. But here, too, it's also important not to go overboard and make the analytical panels neat.
Major mistakes when developing dashboards
Let's summarize with a list of common mistakes when creating a dashboard:
- Ignoring user needs. The analytics stage is mandatory. Without it, it's impossible to understand what should be on the dashboard. As a result, you can spend a lot of money on a fancy tool and find out that no one uses it.
- Place too much data. Give the user only the numbers they really need so as not to take up time with unnecessary data. Split the dashboard into multiple screens if there is a lot of information.
- Use complex graphs and charts. The user needs to scan the visual, so focus on standard patterns to not confuse the user.
- Place data close together. Don't try to fill every pixel. Otherwise, instead of a structured screen, you get a canvas of information that's hard to navigate.
- Make the dashboard monotonous. Without accents in the form of colors, size of elements, and fonts, there will be nothing to catch the eye, and the user will spend more time studying the data.
- Use a monotonous palette. Close colors and shades will merge, so you lose an additional tool for communication with the user interface.
The main sin in developing a dashboard is to score on the main purpose. It should help you make decisions and speed up the analytics process. A bad dashboard wastes the user's time and budget, while a good one saves hundreds of person-hours.
MaybeWorks experience in dashboards development
One of the last dashboard development projects was an instant annotation & screen capture tool for fast visual feedback on live web pages, PDFs, and more.
Our Full Stack developers augmented the client's team to improve and extend existing dashboards. They were implementing new features, extending functionality, doing bug fixes, and improving the UI using React, TS, Firebase, and Recoil. Completed tasks:
- Code optimization and refactoring
- Web tools implementation
- Resize implementation
- Share functionality
Feel free to contact us today to get your team augmented with MaybeWorks developers.