AI Persona Tiers
Every AI persona in atypica.AI belongs to one of four tiers. The tier tells you how rich the persona is and what you can do with it — who else can find it, whether you can edit it, and how it's typically used.
Most of the time you don't pick a tier directly. Scout, interviews, and persona import all produce a tier automatically based on how the persona was built. But you do need to know what the tier means when you see one in the UI or pick a persona for a study.
The 7 dimensions
All personas are built around the same framework, regardless of tier:
- Demographics — age, gender, income, education, occupation
- Geographic — city, neighborhood type, climate
- Psychological — values, lifestyle, motivations, fears
- Behavioral — habits, purchase patterns, decision process
- Needs & pain points — unmet needs, frustrations, workarounds
- Tech acceptance — openness to new products and digital tools
- Social relations — family, friend group, community belonging
How many of these dimensions a persona covers is the main thing that determines its tier.
The four tiers
Tier 0 — Default synthetic. Generated on the fly when a persona is needed and nothing better is available. Used for prototyping, throwaway scenarios, or behind-the-scenes scaffolding. You usually don't interact with Tier 0 directly — it shows up only when nothing else fits.
Tier 1 — Standard library persona. Covers most of the basic dimensions. These are the workhorses in the shared library — usable for concept testing, trend exploration, group discussions, and most research questions.
Tier 2 — High-fidelity persona. Covers all or nearly all of the 7 dimensions, often built from deep observation (Scout) or from interview-derived data. Use Tier 2 when the decision is high-stakes or the audience is hard to reach. Tier 2 is what Scout produces when it builds personas from observation.
Tier 3 — Your imported persona. Personas you import from CRM data, past interviews, or first-party research. The persona is private to you — no one else can find it in the shared library, even if the content is similar. Tier 3 is about ownership, not quality.
What tier affects in practice
| Tier 0 | Tier 1 | Tier 2 | Tier 3 | |
|---|---|---|---|---|
| Appears in shared library search | ❌ | ✅ | ✅ | ❌ (private) |
| Usable in Interview / Discussion | ✅ | ✅ | ✅ | ✅ |
| Can be edited | ❌ | ❌ | ❌ | ✅ |
| Default for Scout-built personas | rarely | sometimes | usually | n/a |
The practical effect:
- Tier 1 and Tier 2 are the "shared library" — anyone can search and use them.
- Tier 3 is your own private space — use it for customer data you don't want in the shared library.
- Tier 0 is internal scaffolding. You don't usually see it unless something went wrong with persona generation.
When the tier matters
The tier matters most when:
- The decision is high-stakes (new product launch, brand repositioning). A low-tier persona will give you generic, low-signal feedback that may mislead.
- You're testing something subtle (a tone of voice, a specific feature nuance). Low-tier personas won't surface these.
- Privacy matters. If you're researching your own customers, use Tier 3 so their data doesn't appear in shared search results where other users could find it.
The tier matters less when:
- You're exploring, not deciding. Tier 1 is fine for "what should we even ask?"
- The audience is broad and the question is broad. Tier 1 spot-checks are enough.
How tier gets assigned
You don't usually pick. Tier is automatic:
- Scout running on social observation → usually Tier 2.
- You import your own data through the persona import flow → Tier 3 (always private).
- Interviews enriching a persona over time → can lift a Tier 1 to Tier 2 as more dimensions get covered.
- Default / fallback persona generation → Tier 0.
The system tracks dimension coverage as the persona is built or refined. A persona that gains more dimensions over time can move up a tier.
Library size
The shared library has grown into the hundreds of thousands of personas. There's no hard cap — personas accumulate as users (and Scout) build them.
What this means for your studies
When you start a Discussion or Interview, you pick personas from a search. That search shows you Tier 1 and Tier 2 from the shared library, plus your own Tier 3 personas. You can filter by dimension coverage to find personas that match the depth of decision you're making.
For most research, Tier 1 personas are fine. Reach for Tier 2 when you're making a high-stakes call and want the richest signal available.