As the kick off for Dashboard Week, our mission was to rebuild a dashboard in Tableau Next.
Here’s my original dashboard, built using Real World Fake Data earlier in training.
You can view it on my Tableau Public here.
We first explored Tableau Next yesterday, working through an internal training.
According to Tableau: Tableau Next is a flexible, API-first analytics experience that integrates the #1 analytics platform with Agentforce, the world’s first digital labor platform. Built on a composable architecture, with a unified data layer and trusted semantics, it delivers personalized, contextual, and actionable insights to every user.
For future readers, this blog was written July 30, 2025. There are consistent updates, so hopefully many of the challenges and workarounds that I mention here are solved by the time you are building in Tableau Next yourself.
My general understanding is that Tableau Next is meant to be integrated into the larger Salesforce ecosystem, so less intended for experienced Tableau users and more for folks who want to really utilize Agentforce AI question asking capabilities.
Initially, I tried to go visualization by visualization to rebuild, but I quickly realized that wasn’t appropriate or feasible. Instead, I decided to utilize metrics (similar to Tableau Pulse metrics) for KPI cards, and to reconsider the purpose and flow of the dashboard. Because of limitations (mentioned below in features I’d love to see implemented) regarding parameters, this Tableau Next dashboard couldn’t provide the opportunities for interaction and deep dives that the original dashboard did, so I had to pick one measure and stick to it.
I ended up keeping the same general breakdown of the dashboard- first looking at KPIs/metrics, and overall sales by state, and then looking more into the rewards membership program. One of the challenges of the initial dashboard build was that the data is largely trendless, so it was an opportunity to look at what insights could be useful, rather than what actually is in the existing data.
Some features I’d love to see implemented:
- The ability to adjust data type later than when you create the semantic layer.
- Parameters- while I was able to build a parameter and a calculated field that would be used to swap between measures, the labels on charts and in tooltips were not dynamic.
- The ability to edit or delete no longer needed metrics and calculations. Not being able to edit metrics meant that there is now a graveyard of similarly named metrics as we figured out
Here’s a screenshot of the error message I received when I tried to delete an existing field:
Scary!
- A little more consistency in feature availability. For some reason, two people in our cohort were able to build tables using the Suggest button, but that wasn’t showing up for the rest of us. This is a separate way to build a table than the one we were originally shown during training that was also not showing up for us.
- I’d love to be able to show sheet titles.
- In a similar vein, the ability to hide field headers and legends.
- Increased formatting capabilities in general. I know this is supposed to be a grab and go tool, and the stock options are solid, but
Quick little tips:
- If your labels aren’t showing up, make the viz fit entire view, and they might decide to show up.
- If the viz you put into the dashboard doesn’t match up with the version you have saved, just do a full refresh of your browser.
- In titles, you can’t currently align up and down, but you can throw an extra line in above the text to shift it down some.
My general take away from this experience is that it will be great as a stand alone visualization tool for basic insights, but if you’re used to the customization capabilities of Tableau Desktop it may be a challenging transition.
It’s also a very helpful reminder to approach a dashboard build with the purpose and audience in mind, and then use the tool you have to make useful insights clear.
Without further ado, here’s the dashboard built in Tableau Next.
Here’s to hoping tomorrow’s project requires a little less patience!