A traditional psychometric assessment tells you what the candidate is like. That sounds useful until you realize it doesn’t answer the question that matters.
The question that matters is not “what is this person like?” It is “will this person work well in this role, at this company, in this country, at this moment?”.
Those are different questions. And most standardized assessments dominating the market are designed to answer the first, not the second. This is why so many companies implement an assessment, see it “work” in demo, and 12 months later discover that its correlation with real performance is disappointing.
The structural problem of standardized assessments
Take a typical psychometric assessment from the market. The promise usually sounds like: “we measure N personality/behavioral dimensions against a normative base of millions of professionals.”
Sounds good. But there are three structural problems with that approach:
Problem 1: The normative base isn’t your population
The most widespread assessments are normed against predominantly corporate, Anglo, urban populations. When you apply them at a manufacturing SME in Rosario, a fintech in Bogotá, or a BPO in Lima, the “norm” the candidate is compared against doesn’t represent the real operational context.
It’s not a problem of psychometric validity (the scales measure what they say they measure). It’s a normative interpretation problem: a “high” score on dimension X in the normative population may be “average” in your real population, and vice versa.
Problem 2: They’re not calibrated against your role
An extraversion score of 75 may be ideal for a Sales Development Rep and catastrophic for a Software Engineer working in async deep work. The standardized assessment hands you the 75 — you have to infer the “ideal per role” yourself.
In practice, almost nobody infers it. The report lands with the hiring manager, they look at the radar, see green-green-yellow-green, and decide. The final decision is still intuition disguised as data.
Problem 3: They don’t consider your specific values
“Culture fit” isn’t OCEAN. OCEAN is personality. Culture is shared values. And values are company-specific. A generic assessment can’t measure fit against values it doesn’t know exist.
Most try to solve this with a “universal values” module (typically derived from the Schwartz model). It helps, but it’s a proxy. If your company has “make things happen” as a value (it’s not theoretical — it’s a real value we see articulated at companies we work with), no standardized assessment is going to catch it.
The alternative: evaluation against your specific context
Talen.to is built on the opposite premise: the assessment doesn’t compare the candidate against a general population, but against a specific role × company IdealProfile. That implies several things worth unpacking.
1. OCEAN extended to 6 dimensions
The classic Big Five measures 5 dimensions (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism/Emotional Stability). Talen.to adds a sixth: SR (Structure & Rhythm).
SR captures how a person operates under operational load and shifting context. It’s particularly relevant in LATAM, where the environment changes quickly and the capacity to keep structure without rigidity (or to adapt pace without losing consistency) is an empirical predictor of performance that the classic Big Five doesn’t capture well.
We didn’t invent the dimension from scratch — it’s based on the literature on conscientiousness facets (Roberts et al., 2005, “The Structure of Conscientiousness”) and empirically validated against real field data.
2. Country-level cultural factors
The scoring engine adjusts raw OCEAN scores by cultural context. This isn’t ideological — it’s psychometrically correct. The cross-cultural psychology literature (Hofstede, McCrae, Triandis) documents systematic differences in the expression of OCEAN traits across cultural contexts.
An extraversion score of 70 in Japan isn’t directly comparable to a score of 70 in Argentina. If your company hires globally, ignoring these adjustments introduces systematic bias against candidates from less extraverted or less self-assertive cultures.
3. Custom value sets per company
Each company defines its own values in Talen.to:
- Value name (free).
- Operational definition (what it means at your company specifically).
- Relative weight (
VALORES_WEIGHTS). - Minimum fit threshold (
VALORES_MIN_FIT, typically 75-85%).
If you don’t have your values formalized, there are AI suggestions to start (based on your industry + size). But the final set is yours, not ours. And it’s versioned — when your values evolve, old assessments remain bound to the definition in effect at the time.
4. The 10 archetypes calibrated with real data
This is, in my opinion, the most underused scientific asset in the product. Talen.to classifies each candidate into one (or a blend) of 10 behavioral archetypes:
| Archetype | Dominant behavioral pattern |
|---|---|
| Connector | High social/extraversion, oriented to linking people and information |
| Analytical | High logic/openness, oriented to understanding before acting |
| Entrepreneur | High energy/autonomy, oriented to initiating and pushing |
| Collaborator | High agreeableness/functional conformity, oriented to the team |
| Mentor | High agreeableness + social capacity, oriented to developing others |
| Specialist | High logic + structure, oriented to technical depth |
| Executor | High structure + pace, oriented to consistent delivery |
| Strategist | High logic + autonomy, oriented to plan design |
| Adapter | Broad balance + openness, oriented to changing contexts |
| Operator | High structure + functional conformity, oriented to stable systems |
The archetypes are derived from 7 traits (autonomy, social capacity, pace, conformity, energy, logic, ingenuity) computed from the OCEAN + Values profile. Classification is by 7D Euclidean distance to the closest centroid (matchScore = (1 - dist/√112) × 100).
The centroids aren’t invented. They’re empirically calibrated against real field data by our people science team. That’s the asset: these aren’t archetypes derived from a desk-bound theoretical model — they’re patterns observed in real data from LATAM professionals.
Detail in /features/archetypes.
Why having a proprietary framework matters more than it seems
The relevant thing isn’t the size of the initial calibration sample. It’s what gets done with that data and how the framework sharpens with every hire:
- Validation that the 7 derived traits have coherent factor structure.
- Identification of real centroids (not arbitrary assignment of “this is being a connector”).
- Calibration of matchScores against observed centroids, not against theoretical ideals.
Compare with the usual approach: a research team defines 10 archetypes in a workshop, writes descriptions that sound good, and the platform assigns candidates to those archetypes without ever empirically validating that those archetypes exist.
The difference is the same as between a map made by exploring the terrain and a map drawn on a whiteboard. Both look like maps. Only one gets you there.
What this means for your hiring
Three practical consequences:
1. Your IdealProfile is yours, not ours
When you create a role in Talen.to, you define (with optional AI suggestions) the ideal profile: optimal OCEAN range, prioritized values, required competencies, preferred archetypes. Each candidate is compared against that profile, not against a generic norm.
2. Results are interpretable without a course
Fit isn’t an opaque number. When a candidate gets 78% overallFit, you can break it down:
- How much comes from OCEAN fit vs. Values fit.
- Which specific dimensions have a gap.
- Which are
criticalGaps(gap > 1.0 in some dimension) vs. tolerable gaps. - Which archetype is dominant and how it aligns (or doesn’t) with the role’s desired archetype.
This gives the hiring manager actionable information: “this candidate has 78%, the gap is in Emotional Stability for a high-pressure role, is it developable or a deal-breaker?”. I go deeper on interpretation in The 7 Most Costly Mistakes When Implementing Culture Fit (mistakes 4 and 6).
3. The engine improves with your usage
After 12 months, when you have hundreds of evaluated candidates and active employees hired via the platform, you can correlate scores with real performance and recalibrate. The engine you use in 2027 is calibrated against your company, not against our initial dataset. This is covered in more detail in Algorithmic transparency in HR Tech.
The honest objection: “isn’t it too configurable?”
Yes, if you configure everything without knowing what you’re doing. No, if you use reasonable defaults at the start and adjust when you have signal.
Our practical recommendation for companies starting out:
- Month 0-3: use defaults. Don’t touch weights. Define your own values (or take the AI suggestion). Start generating data.
- Month 3-6: review the first assessments with hiring managers. Detect if there’s systematically a role where fit underestimates or overestimates. Adjust weights for that role.
- Month 6-12: first calibration with real performance. Run
/calibration/recalculate-allif the data justifies it.
Configurability doesn’t mean “you have to touch everything.” It means “you can touch what you need when you need it.”
Closing
Culture fit measured against a generic population tells you what the person is like. Culture fit measured against your role, your company, your cultural context, and your historical data tells you whether the person will work here.
That second question is the only one that matters when deciding to hire.
If you want to see the rest of the platform inventory: What Talen.to actually does. If you want to try it with a real role: create an account at app.talen.to/sign-up or book a demo.
— Clara
Related OCEAN+ profiles
Discover which personality dimensions to look for in each role.
Related Articles
Talen.to's 10 behavioral archetypes (and why we calibrated them with our own data)
These aren't archetypes borrowed from a management book. They're centroids computed in a 7-dimensional space from real field data. How the model works, what it predicts, and why in-house calibration changes everything.
Algorithmic transparency in HR Tech: why your scoring engine should be auditable
Most assessment tools are black boxes. Why that's bad practice, what regulation requires (EU AI Act, NYC AEDT), and what an auditable, configurable scoring engine looks like.
Multi-evaluator competency assessment: when the psychometric test isn't enough
A psychometric assessment tells you who the person is. A multi-evaluator competency assessment tells you how good they are at what the role actually demands. When you need one, when you need the other, and how they combine.