How Tenezet evaluates sales candidates.

1.Why an assessment, not a gut interview

An hour-long «tell me about yourself» interview isn't bad. It's just very imprecise.

Three decades of research point in one direction: if you hire 10 salespeople from a regular conversation, about 5 of them won't last 18 months. If those same 10 went through a structured assessment with concrete tasks and a role-play, on average 2–3 fall short. Not perfect. But that's +2 successful hires per ten — at a salesperson's salary, that's two paychecks you didn't waste.

Gut feeling doesn't go away. It stays as the last filter. But it works twice as well when there's data in front of it.

2.What we measure — five qualities

A good salesperson isn't one quality. It's five separate things, and they don't substitute for each other. This list isn't ours — it's based on 30 years of research into what separates top sellers from average ones.

1. 1

   Achievement motivation. — How much fuel they have — will they keep chasing results when the quarter gets hard.

2. 2

   Sales adaptability. — The ability to switch approach on the fly when a conversation goes sideways.

3. 3

   Confidence and initiative. — Whether they take charge of the conversation. Whether they close on a concrete next step.

4. 4

   Stress resilience. — Whether they hold form when the client pushes, refuses, or negotiates hard.

5. 5

   Sales knowledge. — Whether they understand how a deal usually flows — open, dig, present, close.

Each quality is scored separately. A candidate who is «average» on all five usually outperforms one who is «excellent» on one and «weak» on four.

3.Scale 1, 2, 3 — no vibes

Each quality gets a score from 1 to 3. But it's not «how it felt to me». Each score has a concrete definition.

• 1

  Behavior typical of a weak salesperson. For example: doesn't take initiative, doesn't uncover needs, backs off at the first «no».

• 2

  Behavior of a typical salesperson. Does the basics, but without spark.

• 3

  Behavior of a strong salesperson. Concrete steps, concrete results, concrete skills.

These aren't school grades. They're like the checklist on a driving exam: not «good driver / bad driver», but a list of specific things that get checked. Did they signal the turn? Check the mirrors? Park inside the lines?

4.AI first, then you

After every AI interview and every role-play, our AI reads the transcript and proposes a first set of scores with quotes from the candidate's answers.

AI doesn't make the decision for you. AI is the lab tech. The lab tech does the blood test. You are the doctor. You see the numbers, read the quotes, watch the role-play with your own eyes, and make the final call.

Why we need a lab tech:

• AI doesn't get tired after an hour of conversation, and it doesn't mix up candidates.

• AI doesn't score higher because the candidate is «pleasant company» — that's the halo effect, where one strong impression colors everything else.

• AI reads the transcript faster than you do. An hour of dialogue = 5 minutes of summary.

You can disagree with AI. You often will — that's normal. The point isn't for AI to be right. The point is to give you a structured first read instead of a blank page.

5.What you'll get

Once a candidate has been through every stage and you've entered your scores, we produce the final report — a 1–2 page document:

• Verdict. — «Hire», «hire with reservations», «pass» — with one sentence of why.

• Strengths. — What works for this candidate.

• Risks and gaps. — What to watch out for.

• What to verify next. — 2–4 concrete items if the verdict is «with reservations» — what to discuss, recheck, or test.

Behind the report sits all the working data: interview transcripts, scores per competency with quotes, observations from the role-play. If you want to dig in — everything is there.

6.What Tenezet doesn't do

So there's no confusion between us.

• We don't predict the future 100%. Nobody does. The best assessment methods get it right roughly 7 out of 10 times. Much better than a guess, but not perfect.
• We don't remove all biases. We remove some — the ones that depend on mood and fatigue. But AI is trained on data that can contain its own skews. That's why the decision stays with you.
• We don't consult on hiring. We give you data. The decision, the compensation policy, the offer negotiation — those are your job.
• We don't show candidates their scores. Candidates don't see their scores and don't get feedback on how they compare to others. That's an internal decision for you.
• We don't work without you. AI does the first read. The final call is yours. If you want to delegate the decision to AI — this isn't the product for you.
• We don't assess complex B2B sales. Long cycles, enterprise deals, account management. Our methodology was calibrated on short cycles — the rep closes the deal in a day to two weeks. Long cycles need different predictors that we haven't validated.
• Tour-operator / travel sales — work in progress. Direct meta-analytic evidence for this niche doesn't exist yet. We're generalizing from adjacent service-occupation findings. If you're hiring in tourism, the methodology should work — but we're still gathering criterion data to confirm it.