A CV tells you what they say they can do. This shows you.
Short skill assessments, sent before the interview, with every question tied to a specific skill in your competency framework. Results come back as a level per skill — not a percentage, not a pass mark — on the same scale you already use for hiring and for reviews.
Your campaign had 12,000 sessions and 240 form submissions. Which metric correctly reflects the campaign’s conversion performance?
Answers are saved as they go. Every candidate gets their own results, whether or not they’re selected.
- Tanca HRM
- Skillify
- Vingroup Retail
- Masan Consumer
- GrabFood VN
- Golden Gate
Everything on a CV was written by the person you’re evaluating.
“Proficient in Excel” appears on roughly every CV you receive, written by people whose actual ability ranges from pivot tables to opening the file. You find out which in week three, after the interview slots, the offer, and the first payroll run.
Twenty minutes of scenario questions, before the interview, tells you what a CV structurally cannot — because the candidate is doing the thing rather than describing it.
Unlimited assessments. The neighbouring module is the one with an allowance.
Assessment platforms usually charge per test sent. This one doesn’t, at all.
Unlimited, permanently
Draws on your AI allowance
Written assessments cost us almost nothing to run, so they’re free without a ceiling.
See AI Interview →And the candidate is never charged for anything. Not to take it, not to retake it, not to see their own results.
“What does GA4 stand for” tells you nothing.
Recall questions measure whether someone revised. Scenario questions measure whether they can work something out — which is the thing you’re actually hiring for.
Every question in the bank puts the candidate in a situation with real numbers and asks for a judgement. A store supervisor gets a shift that’s short-staffed on a Saturday. A warehouse lead gets a delivery window that no longer fits. A marketer gets 12,000 sessions and 240 form submissions.
Each question is tagged to one skill in your competency framework, so the result is a level on that skill — not a percentage of a test.
Your Saturday shift is short-staffed by 2 people and the lunch rush starts in 45 minutes. What do you do first?
Level 3 · Store operationsA delivery window has been reduced from 4 hours to 90 minutes due to a supplier delay. The load is 340 items. What is your first action?
Level 3 · Logistics planningYour campaign had 12,000 sessions and 240 form submissions. Which metric correctly reflects conversion performance?
Level 3 · Marketing analyticsPick the skills. The questions are already there.
Or write your own — the bank is a starting point, not a limit.
Choose the framework.
The assessment inherits the skills and weights you already defined for that job title.
Pick the skills to test.
Not all of them — the three or four you can’t tell from a CV.
Set the length.
Fifteen questions, around twenty minutes. Longer tests don’t get finished.
Send it.
From the pipeline, to one candidate or to a whole stage at once.
Write your own questions any time.
Tag each one to a skill and a level, and it joins your own bank for every future assessment.
One question at a time, no tricks, no countdown pressure.
A test designed to make people anxious measures anxiety. This one is designed to measure the skill.
Taking it
Answers save as they go. Close the tab, come back tomorrow, pick up where you stopped.
A Saturday shift is short-staffed by 2 people and the lunch rush begins in 45 minutes. What do you do first?
Their results
Level per skill, on the same scale, same day, whatever you decide. No overall percentage. No comparison to other candidates.
What they’re told up front
The three things they need to know, before the first question. Not buried afterwards.
Most assessments don’t get finished. Here’s the mechanism.
A candidate is being asked for twenty minutes of unpaid work by a company that may never reply to them. Half of them decide it isn’t worth it, and they’re not wrong.
The usual trade
Here
This is an operational argument, not an ethical one.
People finish tests that give them something. Returning the results costs you nothing, takes no extra step, and it’s the difference between assessing most of your shortlist and assessing half of it. It also happens to be the decent thing to do, but you don’t have to care about that for it to work.
A level per skill, not a percentage.
A score of 78% tells you nothing you can act on. A result that says level 4 on client relationships, level 2 on inventory control, against a role that needs level 3 on both, tells you what to ask about in the interview.
Because every question is tagged to a framework skill, results land on the same five-level scale your framework, your AI scores and your performance reviews already use.
Compare candidates for the same role on the same skills, side by side, without anyone doing arithmetic.
What we won’t do to your candidates.
Some of these would make us money. They’re published anyway, because an assessment platform that sells advantages to anxious people produces results you can’t trust.
A candidate never pays for anything that changes the data being scored.
No paid retakes, no paying to finish a question the timer cut off, no paying to re-record an answer. When a competitor sells that, the people who paid submitted a fuller answer set than the people who didn’t — and the results stop being comparable.
The same rubric for everyone.
Every candidate for a role gets the same questions tagged to the same skills, marked the same way. No variation by who they are or what they paid.
Practice is completely separate.
Candidates can practise as much as they like in their own area. None of it is recorded in their profile, sent to you, or counted toward any application — and they’re told that before they start.
No automatic decisions.
The assessment returns levels. It never marks anyone as not continuing, never removes anyone from your view, and never makes a decision.
The assessment is the first time a candidate becomes more than a document.
Up to this point, a candidate is a file someone uploaded. After twenty minutes of assessment, they have verified skills with evidence behind them — the thing a CV can never contain.
It’s also the moment they’re asked, once, in plain words, whether other companies in the system may see their competency profile when a matching role opens. Not buried in a terms page, and off unless they turn it on.
They say yes at a meaningful rate here, because this is the point where they’ve just received something real: their own results, free, with nothing attached.
One scale, from the test to the review.
Competency Framework
Defines the skills and levels
Assessments
You are here
Pipeline & Candidates
Results land on the candidate record
Talent Pool
Verified skills make people findable later
Skillify
Learning paths for the levels not yet reached
The level someone reached in an assessment is directly comparable to the level they reach in their review six months later — because it’s the same scale, defined once.
Measured inside the product.
Chuỗi bán lẻ thực phẩm · 22 cửa hàng
Assessment completion rate after returning results to candidates
Công ty logistics · 3 kho miền Nam
Time to produce a shortlist from submission close to interview invite
Trung tâm tuyển sinh · telesales
Share of new hires meeting target level at their six-month review
You’ll see the same numbers for your own assessments from the first round.
Questions people ask about assessments.
Is it really unlimited?
Yes. Unlimited assessments, unlimited sends, unlimited candidates, no per-test charge. AI Interview is a separate module and carries the monthly AI allowance.
How long should an assessment be?
Around fifteen questions and twenty minutes. Longer ones get abandoned, and an abandoned assessment tells you nothing.
Is there a pass mark?
No, deliberately. Results come back as a level per skill against your framework. Whether that’s enough is your call, made by a person.
Can candidates retake?
If you allow it. They are never charged for it — that’s a published commitment, not a current pricing choice.
How do you stop people looking up answers?
Scenario questions with role-specific numbers are hard to search for, because the answer requires working with those numbers rather than recalling a fact.
What do candidates get?
Their own levels per skill, the gap to the role’s requirements, and a learning path if they want one. Free, same day, whatever you decide.
Can we write our own questions?
Yes. Tag each to a skill and a level and it joins your own bank.
Does this need a competency framework first?
It works best with one, because that’s what turns results into levels. You can run an assessment without one, but you’ll get raw results rather than levels.
Find out before the interview, not after the hire.
Build your first assessment from your framework in about ten minutes. Send as many as you like.
Start free