Editing Fields and Dataset Metadata

Open a lens on a freshly built dataset and it looks... rough. Dimensions are called AccountId.Name and StageName. Amounts show as bare numbers with no currency symbol and eight decimal places. Every chart of your stages picks different colors than the last one.

All of that is dataset metadata, and you fix it from inside a lens — via the Fields panel.

You need the Manage Analytics or Edit Dataflows permission to make these edits. Without it, the Fields panel is read-only.

Opening the Fields panel

From an open lens, look for the Fields option (usually in the lens menu or alongside the field list). It lists every field in the dataset, split into Dimensions, Measures, and Dates. Select a field and its settings appear on the right.

Four things are worth your time.

1. Default fields for the values table

When someone opens this dataset and switches to a values table, which columns should they see first?

Left alone, you get whatever order the dataset was built in — often IDs and system fields at the front. In the Fields panel you can set the default fields: pick the handful of columns that actually tell the story (Account Name, Stage, Close Date, Amount, Owner) and set their order.

It's a small change with an outsized effect. Every colleague who opens that dataset now starts from something readable instead of a wall of eighteen-character IDs.

2. Relabelling dimensions

Select a dimension and edit its label. AccountId.Name becomes Account Name. StageName becomes Stage. Owner.UserRole.Name becomes Sales Role.

This is the highest-value, lowest-effort thing in this lesson. The label is what appears in field pickers, chart axes, legends, tooltips, and table headers — everywhere a business user reads. Nobody should have to know your object model to use your dashboard.

3. Editing values and assigning colors

Expand a dimension and you can see its values, and against each one set a color.

This is how you get consistent, meaningful color across every chart in your org. Pick a color for each stage once — Closed Won green, Closed Lost red, everything in between on a sensible ramp — and from then on every chart grouped by Stage, in every dashboard, uses those colors automatically. No more re-picking colors per chart, and no more "wait, is green won or lost on this one?"

Do it for the dimensions that appear everywhere: Stage, Type, Industry, Priority, Status. Red/amber/green for anything with an inherent good-to-bad ordering pays for itself immediately.

You can also edit the displayed value itself — but here's where you need to be careful.

4. Measure number formats

Select a measure and you can set how its numbers render across the whole platform:

The currency step catches people out: rather than a long dropdown of world currencies, you typically choose the currency or custom format option and then paste the symbol into the format field. Copy the character you want, paste it into the prefix, done. A rupee or euro symbol behaves exactly like the dollar default once it's in there.

Saving

Save, and the changes are live. Not "live in this lens" — live in the dataset. Reopen any existing dashboard built on it and the labels, colors, and currency formats are already applied, with nothing to rebuild.

Do this once per dataset, early, and every dashboard built on top of it starts out looking finished. Next, we let the platform do some of the querying for you: Conversational Analytics, where you ask for the chart in plain English.