Step 4:Charts and dashboards

Create interactive dashboards and visualizations

Viewers

BI specialist

Administration

Introduction

In DataOwn we use Lightdash for creating charts and dashboards. As a user, you work within one Lightdash project specifically set up for your organization. The data models with the correct measures (metrics) and dimensions are set up by you or your colleagues as described in the "Create models" chapter. A limitation is that you currently need to log into LightDash once and then do a "DBT run" to see the models.

You can save charts (graphs and tables) within a "space" or only within a dashboard. The former can be seen directly by end users and you can also use them in multiple dashboards. However, it can quickly become unclear, so we recommend creating charts and tables by default within a dashboard and also working from a dashboard. That means you first create your dashboard and then create the required charts and tables via "Add tile" / "New chart".

For a comprehensive introduction to Lightdash, you can consult the official documentation: Lightdash: Learning to explore data

Creating a dashboard: step by step

  1. Create new dashboard - Click on 'New' in the navigation bar at the top and select 'Dashboard'. Give your dashboard a name and optionally a description, and click on 'Create'. You are automatically redirected to your empty dashboard.

  2. Add tiles - Click on the large 'Add tile' button in the middle of your dashboard. You have two options:

    • 'New chart' - Create a new chart exclusive to this dashboard
    • 'Saved chart' - Add a previously saved chart (this can be reused in multiple dashboards)

  3. Position and resize tiles - Drag tiles to the desired position by grabbing them. Drag the corners or edges to make tiles larger or smaller. Lightdash uses a grid system that automatically shifts other tiles.

  4. Add Markdown or other content (optional) - Click on 'Add tile' and choose 'Markdown' to add text, headings, or explanations to your dashboard. This is useful for context or instructions. You also have the option to create multiple tabs. There is also the option to add Loom video tiles: this is an even more powerful method to give instructions, but Loom is an American company.

  5. Add dashboard-wide filters - Click on 'Add filter' at the top of the dashboard, select the field you want to filter on (e.g. date or region) and choose the filter type (dropdown, datepicker, etc.). Users see the filters at the top of the dashboard and can adjust them without editing the dashboard.

  6. Save dashboard - Click 'Save dashboard' at the top right to save your changes. The dashboard is now available to your team.

Note the difference between edit mode versus view mode. In edit mode, for example, you cannot apply filters.

Tips for a good dashboard

Use tabs

For large dashboards you can use tabs to organize content. While editing, click on 'Add tile' and then '+ Add tab' at the top. Give each tab a name (e.g. 'Overview', 'Details', 'Trends'). Drag tiles to the desired tab.

Less is more

Try not to add everything to a dashboard, it is more important that dashboards look clear and beautiful. As a data specialist or dashboard designer, you can usually handle more clutter than the (other) end users.

Align tiles and ensure white space

Make sure that tiles are precisely aligned horizontally and vertically. Also ensure a white band around and between the tiles. Both of these provide a lot of visual rest.

Start with most important indicators, with standards

The most important indicators (KPIs) belong on the first tab at the top. Here too, you recognize the true master by limitation: it is difficult to manage more than 8 KPIs. These 4-8 KPIs can be elaborated in other tabs, for example one tab per KPI.

Test with end users

Ask for feedback before you finalize the dashboard. Preferably do not only do this with the person who is put forward by the department and is often the most beta. Also test with the perhaps somewhat more alphas in the rest of the department.

Creating a chart: step by step

Creating a chart in Lightdash is simple. Follow these steps:

  1. Open a table - Go to the main menu and select a table (Explore) from the list. These tables contain the data you can report on.

  2. Select your metrics - On the left side you see a list of metrics (Metrics) and dimensions (Dimensions). Check the metrics you want to visualize, for example 'Total revenue' or 'Number of customers'.

  3. Choose dimensions - Select dimensions to group your data, such as 'Month' or 'Product category'. These determine how your data is split.

  4. Add filters (optional) - Click on the filter icon to show only specific data, for example only data from 2024 or a specific region.

  5. Run the query - Click run query to see the preliminary data and check it.

  6. Choose your visualization - In the chart section you see the chart selector at the top right. Click on it to choose from bar charts, line charts, pies, tables, etc. You can also adjust other settings here, such as colours and axis titles. Lightdash automatically adjusts the configuration.

Once you are satisfied with your chart, click the 'Save chart' button at the top right. Give your chart a clear name. Then choose whether to add it to an (existing) dashboard, or which folder (Space) to save it in. This saved chart can be reused later in dashboards and is also visible to end users. In the small variant, Spaces are only for your own overview; in other variants of DataOwn, permissions can be assigned per Space.

Tips for charts

Custom sorting

In Lightdash it is not possible to give a column a custom sort order. However, you can add two dimensions to your model, include both in your dataset, and sort by the one you do not display. For example, if you have months as "jan" through "dec", create a second dimension with 1 through 12. Add both to your dataset, sort by 1 through 12, and display jan through dec in the chart.

Important buttons and UI elements

While working with charts, these are the most important buttons:

  • 'Run query' - Top left, to execute your selection and see results
  • 'Save chart' - Top right, to save your chart
  • 'Configure' (within chart tab) - Top right in the configuration, to change your visualization type
  • Filter icon - In the dimension list, to add filters
  • '...'-menu - On each chart, for extra options such as export or duplicate

Filters are easy to add: click on the filter icon (funnel shape) next to a dimension in the left sidebar. You can then choose from different filter options such as 'equals', 'contains', 'between', etc. Filters work immediately and are visible at the top of your results.

Available visualization types

Via the chart selector at the top right you can choose from 10 different visualization types:

  • Bar charts (Bar/Column) - usually the best choice for comparing categories
  • Line charts (Line) - the first choice for trends over time
  • Tables (Table) - for detailed data overviews
  • KPI tiles (Big Number) - for prominently displaying one important value
  • Gauge - for prominently displaying one percentage

Feel free to experiment with different visualizations - Lightdash automatically adjusts the configuration and you see the result immediately.

In addition to the standard chart types, Lightdash also offers the ability to create custom charts using Vega-Lite. This opens up a world of possibilities for chart types, and also full control over how it looks. However, there are some limitations compared to standard charts:

  • Drill-down functionality is not available
  • You cannot click through to underlying data
  • Cross-filtering between dashboard tiles is not supported

For examples of custom charts and configurations, see: Lightdash: Custom Charts Examples

Generative AI tools such as Mistral AI are very suitable to help with creating and adjusting Vega-Lite configurations. However, be careful with what information you give to the model.

Using dashboard as end user

As a user of a dashboard (without edit rights) you can:

  • Use filters - Adjust filters at the top to view different data
  • Export - Click on the menu (three dots) on a chart to export
  • Refresh - Click the refresh button at the top right for the latest data. This of course only makes sense if the data has been refreshed.
  • Share - Copy the URL to share the dashboard with colleagues

You can adjust filters without changing the dashboard itself - ideal for exploration!

Sharing dashboard

Sharing a dashboard is simple: copy the URL from your browser or click the 🔗 button at the top right in the dashboard. Everyone in your organization with access to the Lightdash project can open the link.

In DataOwn, each team works within one Lightdash project. In the small version, all saved dashboards and charts are available to colleagues with access to DataOwn.