Our Dashboard Design Process
We are a Google Data Studio dashboard agency, and we are working on this brief technique while developing dashboards.
We'll go over this method in detail so you can replicate it for your own internal communication.
Our procedure may be broken down into four steps:
- Show Examples
Step 1: Understand
Many times, we took the incorrect turn. We put in a lot of time and effort, but we didn't go anywhere. That's why we spend so much effort figuring out why you're using the dashboard in the first place.
What is the problem?
Every dashboard should answer at least one critical business problem. Otherwise, we may construct another dashboard that generates additional data that will not benefit the business.
Who will use it?
Assume your dashboard users are the design team. What are you going to put on the dashboard? Do you prefer to include a lot of filters and metrics? I don't think so. As a result, we aim to understand the end-users in order to develop a dashboard that they can use effortlessly.
On the other hand, we should supply Google Ads or Klaviyo professionals with additional analysis features via the dashboard. These dashboards should also include terms that these professionals are comfortable with.
The dashboard should not be either sophisticated or overly basic, but should be well-suited to the users' abilities.
Step 2: Show Examples
In this step, everyone offers to design a wireframe, but our technique is far more practical for the fast-paced commercial environment.
We have a lot of examples and templates from prior projects. We show some of them based on your business requirements and then ask the following questions:
- Do you think it's good?
- What other charts/data do you want to see on this dashboard?
- Do you believe that any of these charts are useless? Which one is it?
Following these questions, both you and we will have a rough idea of the output, allowing for easier/faster progress and smother communication throughout the rest of the process.
Honestly, we tried the theoretical perfect methodology but, we hit the wall because:
- There is basically no possibility that we will comprehend all of your needs on the first meeting.
- Probably, you will have a different understanding when you see the dashboard.
As a result, the dashboard will grow over time, thus there is no need to spend months creating a massive dashboard that would be useless. We prefer to offer the dashboard in weeks so that we may collaborate on the design to create the finest dashboard for your team.
Step 3: Build
This step will largely be on our side. But first, let's take a quick look.
We are spending one-third of our time working on a data connection. If the data is ready and accurate, the rest will be a joy for us because we will be able to see what we have done in real-time.
Here are some data connectors we are working with:
We can also collaborate with various third-party connections. However, in order to remain practical, we try to keep the number of tools to a minimum.
Dashboard Interface Modification
If you like one of our previous works, we will copy it and remove the company-specific parts of the dashboard to provide you with a lightweight starting point.
There are some derivative pages that contain nearly the same content but from a different angle. For instance, we want to make the same page with multiple time spans, such as the last 14 days and the previous 12 months. These two pages are similar yet also distinct. The same since they use the same charts, but different due of changes such as daily/monthly granularity, etc. We're getting rid of the derivatives and only maintaining the core as a starting point.
Because almost every organisation has slightly different wants and issues, we discovered that supplying a template did not solve their problems. That is why we are dedicate time to understanding the specific challenge and the company's analytics maturity.
We now have the brain and skin. Let's work on the muscles in between these.
Data Blending & Accuracy
This stage of the process requires the most thought and try-see loops of any of the steps. We worked with a variety of data sources, but each new instance brought its own set of problems.
So, we should be sure that the data and the blending is accurate in this step.
Work on the Core Dashboard
- Sturdy data connections
- Accurate data blending
- Dashboard core
The first step is to work on the basic dashboard design and appropriately connect the new data sources. As a result, we can see that the infrastructure is operational and that we can continue to improve it.
Implement the Company's Requirements
It is now time to put the dashboard to good use. We will modify the fundamental template to meet your specifications. As a result, the dashboard will provide you with answers to your inquiries.
When you've used a dashboard tool in the past, you could have had the impression that something isn't quite right. This is due to the fact that dashboarding technologies are designed to meet "generic" demands. They are solving common problems, but each firm, including yours, has a particular need to run and optimise the business.
We make every effort to tailor our services to your specific requirements.
Making Effective Charts
Generally, the first iteration will provide the worse user experience. As a result, we wait 1-2 days before presenting the dashboard and take extra care to simplify it. As a result, the first impression will be milder.
Here are some examples of the advantages of going above and above to make your points more powerful and understandable.
- Emphasize the most important data.
- Orient your views for legibility.
- Organize your views.
- Avoid overloading your views.
- Limit the number of colours and shapes in a single view.
Boost the speed
This is the most distinctive aspect of the procedure. You can have the most gorgeous and functional dashboard in the world, but if it is too slow to use, you will not use it, period.
Working with third-party connectors and looking at a larger date range will cause the dashboard to slow down. I don't even want to think about whether or not data blending is practical without cache. You may have to wait several minutes to see the page, and this wait will continue each time you alter a filter/date range, etc. Annoying.
As a result, we cache the data to provide you with the fastest dashboard experience possible.
Step 4: Deliver
The dashboard is already operational, and you can begin reaping its benefits right away. But, while you're enjoying your dashboard, we have work to do.
Testing and Quality Assurance
We will continue to work with precision for one more day. You may also notice certain irregularities while using the dashboard. These will also assist us in locating the issues.
- Accuracy Testing: We need data from the actual source of data in this phrase. For example, if we connect Shopify data then, we need you to export data directly from Shopify's reports and check if there is any discrepancy or not.
- Functionality testing: We always check filters, tables, data selectors etc. before delivering the dashboard but, sometimes we missed some. This final test helps us to fund these small bugs before deliver.
- Design and Consistency Check: We name every page according to its date range and functionality but, sometimes it is same for 2 pages or we put data label on a chart on a page while we forgot for another page. This test will helps us to identify if there is such a problem on the dashboard.
We include a Loom recording with the delivery so you can learn how to utilise the dashboard.
While Training was relatively easy, the adoption of dashboards proved to be more difficult. The dashboard should be used frequently by your team. You know your team better than anyone but, here are some ideas about this phase:
- One method of promotion is the department huddle.
- You can ask questions which this dashboard can answer.
- You can include screenshots from the dashboard on your weekly/monthly update presentation.
Gather feedback and evaulate
For us, this is the most valuable aspect. Because there is a significant difference between a decent and a superb dashboard, despite the fact that the appearance may appear to be similar. Receiving input on the chart from your team will aid us in making the dashboard more "yours."
We occasionally lose focus of the dashboard's objective while developing it and add unnecessary pages/charts to it. This section assists us in retaining what is important on this project.
We can not create a great dashboard on ourselves. We need to collaborate with you to bring the easiest, fastest and most useful dashboard for the target question ever.
We won't create a documentation unless it doesn't contain a BigQuery connection that requires the extensive workforce.
We try to keep ETL process on the Google Dat Studio so, you can play around and see the modification by yourselves.
We are using Notion for project management and the following records will be there naturally and count as a lightweight documentation:
- Purpose of the dashboard
- Your requirements and the revisions
- Questions we asked and your answers
- The first example dashboard we showed you
- Data sources
- Dashboard link
- Walkthrough with a Loom recording
- Bugs and fixes
Tips to Design the Perfect Dashboard
We talked so much about the process. Here are some tips all together :
- Focus on the dashboard users
- Be clear on what you want. Your board's purpose will guide its design
- Select the right type of dashboard and type of charts
- Display the key information at the top of your screen
- Be simple but clear
- Use colours to convey the true understanding in the context
- Make your numbers contextual to help your viewers understand if a particular number is good or bad
- Show hierarchy by using size and position - Make it clear to the viewers what is most important
- Be consistent in everything. Using the same layouts, colours and charts make it easier for users to onboard to the dashboard
- Use interactive elements such as filters and selectors
- You can apply UI design principles here such as using white spaces around every element
- The dashboard ideally should be adaptive for every device but, you can design it for the desktop for cost efficiency
- Continue to improve your dashboard. Make sure your dashboard encourages the right behaviour