Skip to content
Random Truffle
← All agents
02 · Define · Aura

Define
who matters.

Aura turns an opportunity into an audience you can review: inclusion criteria, exclusions, eligibility and the reason each person belongs. Observed behavior stays separate from interpretation and unknowns.

Forty-eight synthetic profiles: 28 eligible, 12 excluded, 8 unresolved. No demographic or sensitive traits inferred.
Guiding question
“Who are we really talking to?”
How it works

Curious and human, attentive to behavior. Aura asks what people did, not just which label they carry, and keeps the unknowns visible.

The problem it solves
Segments get built from a hunch or a demographic label, and nobody can say later who was included, who was left out or why.
What it produces
An audience you can review: inclusion criteria, exclusions, eligibility by channel and the reason each rule exists.

What Aura can do.

  • Build audiences from behavior. First-party data, behavior and context tied to the decision.
  • Define eligibility and exclusions. Supported identity links, channel permissions and suppression rules.
  • Version the audience. Revise the definition as behavior and results change.
Example conversationTelecom · plan upgrade
  1. You

    Define an audience of eligible customers who have not upgraded.

  2. Aura

    I’ll use the eligibility rules and exclude recent upgrades and customers who can’t receive this communication. Review the criteria before I prepare the audience.

  3. You

    Why exclude customers with an open support ticket?

  4. Aura

    An upgrade offer during an unresolved issue tends to read as tone-deaf. It’s a judgment call: keep it or drop it, and I’ll record the decision with the audience.

Why it holds: Rules come from your eligibility and consent data. Customers without a confirmed match stay unresolved; they are not guessed.

OutputAudience logic · for review
RuleCriterion
IncludePostpaid plan eligible for upgrade
Include12 or more months of tenure
ExcludeUpgraded in the last 90 days
ExcludeNo consent for this channel
ExcludeOpen support ticket (to confirm)
NextPrism creates messages for this audience.

Illustrative script with fictional data. It shows how the work feels; it is not a transcript or a product screenshot.

Try Aura in a simulation.

Interactive simulation · illustrative data · not a product screenshot

Who should we speak to next?
Start with the people behind the next decision: inspect behavior, value, intent and affinities in a shared context. This profile can change as evidence changes.
AURA / VISUAL ANSWER

A person-shaped understanding.

Audience portrait
Audience 01

City explorers

IdentityConsented first-time visitors
BehaviorExplores style + running content
ContextCity routines and self-expression
ValueVersatility and discovery
IntentRevisits products before choosing
AffinityFashion, movement, culture
A working hypothesis, not a demographic shortcut. No sensitive characteristics are inferred.

Context: City explorers · creative A

Continue with Prism →

One example,
five stages.

A retailer asks why repeat purchases fell in the North region. The same question, region and period travel through every stage. Illustrative workflow; the numbers are not client results.

  1. 01 · UnderstandNexusRepeat purchases fell 11.8%, concentrated among customers whose last order was 35–60 days ago.
  2. 02 · DefineAura42,380 customers match. Prioritize frequent buyers with no purchase in 45+ days.
  3. 03 · CreatePrismThree retention concepts from your brand rules and product catalog.
  4. 04 · ActivateRadianMeta customer audience, push and email. 80% test, 20% holdout. Ready for review.
  5. 05 · LearnShared systemIllustrative result: 7.4% relative lift in repurchase rate versus holdout. Back to Understand.↺ Returns to Understand