A one-off export is a snapshot. It goes stale the moment you close the spreadsheet, and you rebuild it by hand every time you want a fresh view. A dashboard built on a Relato dataset stays current. You store your numbers in one structured table, add rows as new data comes in, and read the same widgets whenever you want. This article shows you how to set that up.
Create the dataset and its columns
A dataset is a structured table you create in Relato to store data. Start by creating one for your content metrics, then define the columns you want to track. Give each row a date so you can chart change over time, and add a column for every number you care about.
Keep the columns simple and consistent. A row per post per day, or a row per day with totals, both work. Pick one shape and stick to it so your widgets stay clean.
Populate the dataset
You can add or update rows yourself, or have an AI Content Agent fill the dataset for you. An agent can pull metrics on a schedule and upsert rows into the dataset using the Data and Visualisation tools and the Reporting tools. Upserting means new dates get added and existing rows get updated, so you avoid duplicates.
To do it manually, add rows as you collect numbers and update earlier rows if the figures change.
To automate it, set up an agent with the Data and Visualisation and Reporting tools, give it a schedule, and let it upsert rows on each run.
Build widgets on the dataset
Widgets visualize a dataset. Create them from the dataset you just built. Three types cover most reporting:
Time-series chart: shows a number moving over time, like impressions by day.
Scorecard: shows a single headline number, like total clicks this month.
Data table: shows the underlying rows so you can read the detail.
Assemble the dashboard
Arrange your widgets into a dashboard view. A common layout puts scorecards at the top for the headline numbers, a chart below for the trend, and a table at the bottom for the detail. Group related widgets so the dashboard reads top to bottom.
Keep it current
Because the widgets read from the dataset, they refresh as new rows land. If an agent updates the dataset on a schedule, your dashboard stays current on its own. Check that the schedule is running and that new rows are arriving, and your view keeps up without a manual rebuild.
What to track
A content performance dataset often includes columns like these:
Date
Impressions
Clicks
Click-through rate
Published count
Channel or source
Tips
Add a date column from the start so time-series charts have an axis to plot against.
Use upsert rather than insert when an agent runs on a schedule, so reruns update rows instead of duplicating them.
Query the dataset or pull summary stats to sanity-check the numbers before you trust a widget.
Start with a few columns you will actually read, and add more once the basics work.
FAQ
Can an agent keep the dataset updated for me?
Yes. An AI Content Agent can pull metrics and upsert rows on a schedule, so the dataset and every widget built on it stay current without manual work.
What widget types can I use?
Three cover most dashboards: a time-series chart for trends, a scorecard for a single headline number, and a data table for the underlying rows.
Related articles
Data and Visualisation tools
Reporting tools
Scheduling agents and setting event triggers
Measure content ROI: read Google Search Console and Google Analytics together
If you need help, message us in the Relato Slack community or email [email protected].