The process
A much more robust way of finding the right candidate.
The work required from your team is the same or less vs. working with a traditional recruiter.
Step by step
The process. In 8 simple steps.
Before you commit anything
A quick call to understand the role you are looking for. We will produce a summary plan of how we’ll locate the candidate for your feedback so we are aligned. These two simple steps will help us locate some star candidates.
- 01 Day 0
Brief call
Twenty minutes on what you expect practically from the person in this role. We will want to clarify what the role means in terms of problems solved, KPIs moved etc.
What it costs you Twenty minutes
- 02 Day 1 to 2
Summary / scorecard
We will prepare a one pager based on what we heard, broadly what we plan to test / measure against and some salary ranges of this profile of candidate.
What it costs you A further brief discussion with us to cover your comments before we get to work!
The search
The best part of a month in which nothing is asked of you. This is the stretch where agencies go quiet, so it is the stretch we write down.
- 03 Week 1 to 2
Sourcing
We approach people directly for this search, and everyone we put forward sits the assessment as part of it. Nobody sits it on spec, so every score you see was earned on the problems your role involves.
- 04 Week 2 to 3
Assessment
Four parts, narrated aloud and screen-recorded, which ensures that responses can’t just be AI generated. The product of this is a short list to you.
After the shortlist
Your own process, better informed, and the first point at which money moves.
- 05 Week 4
Shortlist
A shortlist of people who we feel could actually do the job (and why!).
- Report The score and report
- Video The screen recordingWith confirmation from the candidate that they are happy for this to be shared.
- One page summary The interview guide
You will know more before the first interview than you usually know after the third.
What it costs you Watch what you want to watch, but the summary we’ll have shared will capture the most important parts
- 06 Week 4 to 5
Your interview
You run your own process and we do not sit in. What we send instead is the interview guide which helps with: some suggestions on what to ask each person, where to press, and what a strong answer sounds like, all drawn from their own session. We do not coach candidates on your questions, and we do not chase you for feedback every other day. Ask for a second opinion on someone and you will get a straight one.
What it costs you Same time taken as a normal interview process, however, a lot more will be learned in the sessions.
- 07 On start date
Offer and start
You make the offer directly. We invoice on the day they start, not on the day they sign, because signed is not started. If they never start, there is no fee.
What it costs you 13.5% of first-year salary + the monthly seat fee
- 08 Ongoing
The seat
The seat keeps working after the hire: a free replacement search if they leave, once the seat has been active three months, annual re-testing and re-calibration if desired so the standard holds, and an interview guide for anyone you put in that seat later.
What it costs you See pricing based on various salary bands
Inside step four
The six capabilities, and the four parts that show them.
The capabilities are the standard and they do not change. The four timed parts are how we get at them, and those do.
- 01
Analytical reasoning
Finds the fault, not the symptom.
- 02
Commercial reasoning
Knows which fix pays for itself.
- 03
Prioritisation
Ranks by damage, not by ease.
- 04
Structured thinking
Orders the problem before solving it.
- 05
Communication
Explains it to someone non-technical.
- 06
Composure
Under time, and under observation.
How we get at them · four timed parts
One example prompt from each. The elements scramble between sittings, because knowing a trap is there is not the same as finding it under a timer, out loud, with the follow-ups still to come.
- 01
Diagnostic
Reply rates on the outbound sequence halved three weeks ago. Here are the sends, the enrichment log and the sequence settings. Find out why.
- Forms a hypothesis before touching anything
- Says out loud what they are ruling out, and why
- Checks the boring causes before the interesting ones
- Recognises when the data contradicts the hypothesis
Shows Analytical reasoning · Structured thinking
- 02
Design
A funding announcement should become a personalised first email within the hour. Design the pipeline: what runs where, what a human still checks, and what breaks first.
- Designs for the failure case, not the happy path
- Knows where the enrichment will be wrong and checks it before anything sends
- Puts the one manual check where a bad send costs most
- Considers who maintains it after they leave
Shows Commercial reasoning · Prioritisation
- 03
Data
Three enrichment vendors, one account list, and none of them agree. The VP wants a target list of two hundred by Friday. Build it, and say what you trust.
- Checks which vendor is right before deciding who is in
- Defines the ideal customer before filtering to it
- States what the data does not support
- Gives a list rather than hiding in caveats
Shows Analytical reasoning · Commercial reasoning
- 04
Communication
Explain what you just did to a VP of Sales who does not care what it runs on.
- Leads with the answer, not the method
- Drops the tooling vocabulary without dumbing down
- Is explicit about confidence
- Says what they would do next
Shows Communication
Questions
What clients ask before the first brief.
About the search
- How long does a search actually take?
- About four weeks to shortlist in most cases. Faster when we already have assessed people who match, slower for genuinely rare combinations. We will tell you which of those it is on the brief call rather than after a month of silence.
- What if none of the shortlist is right?
- We go again at no cost. If the second round misses too, that usually means the scorecard was wrong rather than the sourcing, and we will say so and rewrite it with you. You have paid nothing at this point and you keep the recordings.
- Do you work exclusively?
- No, and we would rather you ran other agencies alongside us. If someone else fills the role first you owe us nothing. Exclusivity protects the agency, not the client, and asking for it from a standing start would be asking you to take our word for something you have no reason to take our word for.
- Can you assess people we have already found?
- Yes, $100 a candidate. You send a shortlist from your own pipeline, they sit the full assessment, and you get the same recording, score and interview guide. If you hire one of them there is no placement fee on top.
- Who actually does the work?
- Rareix is small and deliberately narrow. The person on your brief call is the person running your search. There is no account manager to hand you to.
About the assessment
- Why does the candidate have to narrate?
- Because the output alone does not separate people. Two candidates can reach the same answer, one by reasoning and one by pattern-matching something they saw last year. Narration exposes which. It is also the part that makes the recording useful to you. You are watching someone think, not reviewing a finished artefact.
- How is it marked?
- By an AI model, with a human in the loop. The model marks each part against a written standard, and we review the marking before any score reaches you. The standard was written by practising RevOps leaders and GTM engineers, and each part was tested against real operators before it was used to score anyone. Every score is stamped with the version of the standard that produced it. The model keeps the marking consistent in a way a panel of humans is not. The review is there for the moments a model reads wrongly.
- What stops someone cheating?
- The screen recording and the narration do most of it. A candidate reading from a second window is visible, and someone reciting an answer they did not derive tends to fall apart on the follow-up questions. We are honest that a determined cheat with a second machine is not impossible to imagine, which is exactly why the assessment is recorded and why the follow-up questions exist. We have deliberately not built proctoring, keystroke analysis or webcam monitoring: they are invasive, they punish nervous people more than dishonest ones, and they would not catch the failure mode that matters.
- What does the score mean?
- A score on what the role actually needs, part by part, with where they were strong and where they were not. Every score is stamped with the version of the standard that produced it, so if the standard changes, older scores are recomputed against it rather than silently reinterpreted.
- What do you not test?
- Anything a tool vendor certifies. We do not test whether someone can name a Salesforce object or click through a HubSpot workflow builder. That is learnable in a fortnight and it is what every existing certification already measures. We also do not test culture fit, coachability or motivation. Those are real, and they are yours to assess in the interview. The guide we send is written to help you do exactly that.
Start with the brief call. It costs you twenty minutes.
If the search is wrong for us you will know on that call, not a month later.
Book a call