Conversational Analytics

Everything so far has been you clicking your way to a query. Conversational Analytics flips it: you type the question in plain English, and the platform builds the lens for you.

It isn't magic and it won't replace Explorer mode, but for quick questions — and for business users who will never learn where the grouping button is — it's genuinely useful.

Turning it on

It's off by default. An admin enables it in Setup → Analytics → Settings, where you'll find the checkbox for Conversational Analytics (natural-language querying) alongside the other Analytics feature toggles. Save, refresh, and a search/ask box appears on the Analytics home and on datasets.

Phrasing that works

The single most useful habit: lead with the measure, then say "by" and the dimension.

• "amount by industry"
• "count of opportunities by stage"
• "average amount by account owner"

Type "amount by industry" and you get a chart of summed amount, grouped by industry, immediately — the same lens you would have built with three clicks. From there it's a normal lens: you can switch chart types, add filters, save it, or clip it to a dashboard.

Layer on a second dimension the same way — "amount by industry by stage" — and it nests the groupings, producing a stacked or clustered chart. You can also fold in time and filters: "amount by stage this quarter", "count of closed won opportunities by industry".

It respects your metadata

Notice what the generated chart looks like: if you assigned colors to your dimension values in the Fields panel, Conversational Analytics uses them. Currency formats, number formats, and field labels come through too.

That's the point worth taking away. Conversational Analytics doesn't bypass your dataset metadata — it sits on top of it. Time spent curating fields makes every downstream surface better, whether a human built the query or a sentence did.

Where it fits

Use it as an accelerator: ask the question, let it draft the lens, then take over in Explorer mode for the parts that need precision.

That closes out the section. You can now reach Analytics from any Lightning app, organize and share work through apps, explore any dataset fluently, read and edit the SAQL behind it, and curate the metadata everything else inherits. Time to prove it — the graded quiz and a set of interview questions are next.