Rover BI Power Users: Practical Use Cases & Best Practices from Our Latest Master Class
Every ERP has data. The difference is whether your team can use it—quickly, confidently, and without turning every question into a custom report request.
That’s why this month’s Rover Master Class went deeper than “here’s a dashboard.” We focused on Rover BI best practices for people who build and share reports—whether you consider yourself a power user or you’re simply the one everyone comes to when they need answers.
In this session, our team walked through how Rover data is structured, how to build reliable datasets, and how to turn “gnarly” MultiValue data into clean, usable dashboards—fast.
Here’s what we covered—and why it matters.
Why Rover BI Power Skills Matter
Most reporting headaches don’t come from a lack of data—they come from:
- Pulling too much data (slow refreshes, bloated datasets)
- Rebuilding the same visuals over and over
- Wrestling with arrays and nested fields (hello, line items)
- Spending hours on calculations that already exist in Rover
- Creating dashboards that look right… until someone filters by date
Rover BI solves these problems—but only when datasets are built with the right structure and habits.
Start With the Foundation: How Rover Data is Structured
The class started by grounding everyone in how Rover organizes information:
- Everything is a file (Sales Orders, Customers, Inventory, Parts, etc.)
- Files contain fields (SO ID, Cust Code, Ship Via, Order Date…)
- And the “secret weapon” for BI: Correlatives
Correlatives: Your Shortcut to Smarter Reporting
A correlative is Rover’s built-in way of reaching into another file to bring back related data.
Example:
- Sales Order has a
Cust Code - A correlative can automatically pull the Customer Name from the Customer file—without writing calculated logic
Even better: Rover includes powerful correlatives like AR Balance, Order Amount, and Open Amount—so before you build custom calculations, check correlatives first.
That one habit alone saves hours.
Building Datasets the Right Way (Without the Bloat)
Next, we walked through a live dataset build using Rover’s API connection inside BI. Key best practices:
1) Use the Right Routine
- getRecord = pulling a file (like SO)
- getReport = pulling a report (like R13)
2) Use Selection to Control Volume
Selections let you pull only the records you actually need.
In the example:
- Start simple:
SELECT SO SAMPLE 50 - Filter for open orders:
STATUS = "NEW" - Add date control:
DATE GE 25-1-1(greater than or equal to Jan 1, 2025)
This is how you keep datasets lean and dashboards fast.
3) Use Field Names When You Only Need a Few Columns
If your source file is huge, don’t pull everything.
Pull only what you need + correlatives.
That reduces load time and improves performance dramatically.
Arrays vs. Fields: The Moment BI Starts to Click
One of the most valuable parts of the session was the “aha” around structure:
- A field is a single value (like Cust Name)
- An array is a list (like Sales Order line items)
Sales orders often include line items as arrays, which is why reports can get messy fast.
Normalize: Turn Arrays Into Rows
We demonstrated the Normalize flow step to transform:
- 1 sales order with 10 line items
into - 10 clean rows (one per line)
This is the key to building line-level reports that are actually usable.
Cleaning & Enriching Data with Flow Steps
Once the dataset was pulling the right records, we showed the most useful flow steps for real-world reporting:
Remove Fields
Quickly strip out noise—fields you don’t need for reporting.
Field Settings
Set formatting so BI correctly recognizes:
- dates as dates
- amounts as numbers
- fields that should be sortable/filterable
This becomes critical once you start building visuals.
Flatten Nested Objects with Power Scripts (and AI)
After normalizing line items, you often end up with nested JSON-like objects.
Instead of hand-writing 50 lines of JavaScript:
- use a Rover BI AI assistant to generate the Power Script
- iterate if needed (refreshing the assistant context helps)
This is one of the biggest workflow accelerators for modern BI users.
Pull Fields From Another Dataset
Need part descriptions, attributes, or categories that aren’t in your SO dataset?
Bring them in from your Parts dataset using Fields from Another Dataset and match on Part Number.
This is how you build rich operational dashboards without overcomplicating your source query.
Automation: Jobs, Refresh Schedules, and Update Modes
Once the dataset is right, it has to stay right.
We covered how to:
- remove the
samplelimiter once the dataset is ready - create a Job to refresh on a schedule
- use a timestamp for tracking record loads
Replace vs. Upsert (Huge for Scale)
- Replace: best for smaller “current state” datasets (like open orders)
- Upsert: best for large datasets where only new/changed records should update
This is one of the best ways to avoid long refresh times as BI usage grows.
Building Dashboards Faster with Saved Visuals + Magic Visual
To wrap up, we showed how to turn clean datasets into dashboards efficiently.
Build Visuals at the Dataset Level
When you build visuals inside the dataset:
- you can reuse them across multiple dashboards
- you avoid rebuilding charts repeatedly
Magic Visual: Your Dashboard Shortcut
Magic Visual lets you prompt BI with what you want, like:
- “Create a map of sales by state”
- “Show top 10 customers by order amount”
It generates visuals quickly, and you can inspect the underlying transforms to refine them.
Create Multiple Views of One Dataset
A powerful dashboard pattern:
- add the same dataset multiple times
- apply different filters (Month-to-Date, Year-to-Date, etc.)
- use dynamic date keywords (like “year begin” and “year end”)
This keeps your dashboards clean, consistent, and flexible.
Best Practices Recap (What Power Users Do Differently)
If you only take a few habits from this Master Class, make them these:
- Check correlatives before writing calculations
- Use selection + samples while building datasets
- Limit field names to what you actually need
- Use Normalize for line-level reporting
- Clean early with Remove Fields
- Always include Field Settings for formatting
- Use Jobs + Upsert to keep data fresh without reloading the world
- Build visuals once, reuse everywhere
One Final Thought
Rover BI is where “Chaos to Clarity” becomes real.
When your datasets are structured correctly, you stop chasing data and start leading with it—giving warehouse teams, production managers, sales leaders, and executives a shared, trusted source of truth.
