The uxerhub alternative to Vervoe, for product roles
By Sergio Gualda · 6 min read · Updated
Vervoe is a skills-based hiring platform built around immersive tasks — spreadsheets, code, presentations, video answers — graded by AI, with traceable scores that show how a candidate performed. uxerhub does not ask candidates to produce anything. It presents twelve written situations in which no option is correct, and measures which action they reach for and which they rule out. Both reject the CV. They disagree about whether output or judgment is the thing you can actually observe in an hour.
What Vervoe does better
Vervoe watches people do the job, which is the most direct evidence there is. For roles where the work has an observable output in a short sitting — a spreadsheet model, a support reply, a sales pitch, a piece of code — seeing the artefact beats inferring from a choice. Nothing we do substitutes for that.
Their AI grading is also traceable by design: their site is explicit that hiring teams can see how and why a candidate scored what they scored. That is the right instinct, and it is more than most AI scoring offers.
And the format range is wide — spreadsheets, code, presentations, video — across a general assessment library. If you hire for many functions and want candidates producing real work in each, that is a genuinely different product from ours and a good one.
Side by side
Everything in the Vervoe column comes from their own public site. Where they do not publish something, this says so rather than guessing.
| Vervoe | uxerhub | |
|---|---|---|
| What the candidate does | Produces work: tasks, code, presentations, video answers | Chooses. Twelve written situations, marking what they would do first and what they would never do. |
| What grades it | AI grading, with visible reasoning behind each score | Deterministic arithmetic. No language model in the scoring path and no human grader, so the report exists the moment they submit. |
| What AI does to the result | A candidate using AI to produce the work is producing work an AI can produce — which the grading has to account for | There is no artefact to generate. Nothing distinguishes a strong answer by its polish, and visual polish is explicitly not scored. |
| Time on the candidate | Depends on the tasks; producing real work takes real time | About fifteen minutes, once, for every candidate |
| Scope | General assessment library across functions | Six product roles — product designer, product manager, UX researcher, design engineer, front-end engineer and UX writer. Nothing else, deliberately. |
| Weighting | The assessment you build defines what counts | You name up to three capabilities the role cannot do without, and those weight the score |
Which one you should pick
Choose Vervoe if
- The role has an output you can see in an hour, and you want to see it.
- You are hiring across functions, not just product.
- You want candidates demonstrating tools they will use on day one.
- You are comfortable with AI grading and want visibility into how it scored.
Choose uxerhub if
- The role's real work is deciding under ambiguity, where the artefact hides more than it shows.
- You want the same fifteen minutes from every candidate, so the comparison is fair.
- You want no model in the scoring path at all, for auditability or for the candidate's sake.
- You need to compare candidates against one specific job, not against a general standard.
Output or judgment: the real disagreement
The case for output is obvious: you see the work. The case against is that in a two-hour exercise you mostly see how fast someone produces something presentable, and for product roles that correlates with the wrong things — tool fluency, template familiarity, and now access to a good model.
The case for judgment is that the decisions are the job. A product designer's day is not spent making screens; it is spent deciding which screen is worth making, what to cut when the date will not move, and when to push back on a brief that is wrong. None of that shows up in an artefact, because the artefact only records the choice that was made, never the four that were rejected.
So uxerhub shows the four. Every option in every situation is something a sensible professional might do — there is no obviously wrong answer to avoid — and what gets measured is which one they reach for first and which one they refuse outright.
Where AI changed the maths
Both approaches were designed before a candidate could generate a plausible artefact in ninety seconds. Vervoe's response is to grade with visibility into the reasoning; ours is to remove the artefact from the equation entirely.
Neither is a complete answer. A candidate can still think about our situations with a model open in another tab — we record how long each situation takes and withhold results that were answered faster than the situations can be read, but that is a floor, not a guarantee. Anyone claiming their assessment is AI-proof is selling something.
Frequently asked
- Is uxerhub a skills test like Vervoe?
- No, and the distinction is load-bearing. A skills test asks whether someone can do a thing. uxerhub asks which thing they would do — the options are all plausible, so there is nothing to be right about. It is a situational judgment test, scored against the capabilities you said the role needs.
- Does uxerhub use AI to score?
- No. Scoring is deterministic arithmetic with no language model anywhere in the path and no human grader. That is why a report exists the second a candidate submits, and why the same answers always produce the same score.
- Can I use both?
- Yes, and for a senior hire it is a reasonable sequence: uxerhub at the top to decide who is worth an interview, and a work sample later for the two or three who make it. What is not worth doing is a two-hour exercise on eight people to find out that six were never a fit for this particular role.