From insight to action

Discovery tells you what’s true about your customers. This section is about what you do with that. Start with actions: the quickest way to understand what to do about your use case, because they answer the goal and KPIs from the use case definition across your whole base, each led by a computed headline for what the play could be worth. Every action is composed from segment tactics — one per targeted segment, each contributing its slice of the audience, its angle on the play, and the per-segment evidence the play draws on. In map terms, discovery drew the map. Now you plot routes across it and run them.

Your plays can only pull the levers your data built

There’s a causal chain running through the whole product that’s easy to miss: the features engineered during data preparation become the levers your plays can pull. Every action, plan, and segment tactic is expressed in terms of the columns in your customer 360; nothing else exists for it to reason with.

A concrete example: a “win back lapsed high-spenders” play is only buildable if recency and spend measures were part of the survey. If the prepared dataset has a <days since last purchase> and a spend measure, “lapsed” and “high-spender” are expressible, segments can separate along those axes, and the play can name its audience. If those measurements were never taken, no amount of downstream analysis can conjure them. The same chain grounds the numbers: an action’s value headline is computed from measured audience counts and the value columns your data actually carries.

The practical consequence: when a play you want isn’t expressible in the results, the fix is often one step back in data preparation, not a different reading of the segments. Browse feature engineering examples for the kinds of measures worth having, and raise gaps with your Neuralift contact. Routes are only as good as the survey under them.

The working order

Step	Surface	What it answers
1. Actions	Actions tab	The ranked playbook: which plays answer the use case’s goal and KPIs, what is each worth, and does the math hold up?
2. Segment tactics	The audience card of each action	Who’s in the play, how does each segment’s version differ, and how strong is the evidence behind it?
3. Inside a segment	Segment detail in Discovery	Who are the people behind a play — the persona, insights, and defining features of each targeted group?
4. Export & activate	Exports tab and delivery	How do the audiences get to the systems that run the campaign?

You can enter at any step, but actions are the natural front door: one screen that says what to do about the whole use case. From there, unfold an action’s segment tactics when you want the evidence behind a play, jump into Discovery for the people behind it, and export the audiences out of Neuralift to run it.

Note. Actions and segment tactics are AI-generated guidance grounded in your data and use case definition. Your team decides what ships; review before committing budget.

Where to go next

• Start with the plan: Actions.
• Drill into the evidence: Segment tactics.
• Missing a lever? Revisit feature engineering examples.