There’s a problem with references we all know and rarely say out loud: when you ask someone to assess a former colleague, they tend to be kind.
It’s human. Nobody wants to hurt someone they know. So they mark everything well, give top ratings to almost everything, and move on with a clear conscience.
The result is an assessment where everything is excellent. And an assessment where everything is excellent says nothing: if every skill is worth the same, none of them stands out.
“A report made of ‘all good’ helps no one decide. The information isn’t in the average — it’s in the priorities.”
This is exactly the problem StarmyCV’s system is built to solve. And it’s what makes the report we described last week actually mean something. Let’s take it step by step.
1. The stars are limited
The assessor can’t rate everything highly. They have a limited number of stars to spread across the skills, so they’re forced to choose: where the candidate is genuinely strong, and where they’re merely fine.
That small act of forcing a choice changes everything. Look at the difference:
✗ No limit
Leadership · ★★★★★
Communication · ★★★★★
Teamwork · ★★★★★
Adaptability · ★★★★★
Everything stands out = nothing does
✓ With limited stars
Leadership · ★★★★★
Communication · ★★★★☆
Teamwork · ★★★☆☆
Adaptability · ★★☆☆☆
A real ranking appears
On the left there’s no information: it’s a polite “all good”. On the right there’s a real hierarchy of strengths — exactly what a recruiter needs to decide.
But prioritising with stars still leaves one question open: what if that high rating is just as arbitrary, only spread around? That’s why a second rule backs it up.
Top-rated skills have to be justified: when an assessor gives a skill 4 or 5 stars, the system asks for a short comment to back it up. You can’t hand out a high mark for free. The star stops being a loose number and gains a concrete example behind it.
That comment is often the most valuable thing in the report: it doesn’t just say how much the candidate stands out at something, but why, in the words of someone who worked with them.
The full mechanic, step by step
How the stars are distributed, how many are available and how each rating is justified is something we explain in detail in a separate guide.
→ How to distribute the stars correctly
2. The open questions do the pressure-testing
Stars sort. But what really validates a candidate are the open questions the referees answer: free text, not checkboxes. And they aren’t just any questions — they’re aimed at three things at once.
- 1Test the candidate against the job description: if the role calls for team management, the question steers the referee to speak to that, not in the abstract.
- 2Confirm or challenge the declared skills: what the candidate says about themselves is cross-checked against what someone who worked with them says.
- 3Draw out concrete examples of behaviour: not “a good leader”, but the situation where they proved it.
And they’re the same questions for every candidate in the process. That’s what makes comparison possible: you can only put one answer next to another when both answer the same thing. Different questions give you scattered anecdotes; the same ones give you a real comparison.
This is where stars and text come together: the stars give you the quick ranking, the open answers give you the why and the evidence. One without the other only gets you halfway.
3. Not every voice weighs the same
A reference from a direct manager the candidate worked with for two years can’t count the same as one from someone they overlapped with for a month in another team.
So each assessor states the professional relationship they had with the candidate, and that relationship adjusts the weight of their rating. Closeness and the type of bond are part of the calculation, not a decorative detail.
And all of this rests on one foundation: everyone who assesses identifies themselves through LinkedIn. Not a loose name on a form, but someone with a verifiable track record and a relationship to the candidate you can check.
From ratings to a verified badge
When a specific skill of a candidate gathers enough assessments, it stops being an isolated data point and becomes a verified badge: a recognition that sums up, in a single mark, what several people who worked with them confirmed.
⭐Several star ratings
⚖️Weighted by relationship
📊Aggregated per skill
🏅Verified badge
That badge is what the candidate shows on their profile, and what backs up, with evidence, every claim in the report your client receives. The loop closes: the report isn’t an opinion churned out automatically, it’s the summary of many structured, prioritised, weighted and verified ratings.
🔎
Evidence, not impressions
Every rating has an identified person and a verifiable professional relationship behind it.
📐
Genuinely comparable
Same questions, limited stars and consistent weights: candidates can be measured against the same yardstick.
Why this matters
Over the past weeks we’ve shown what StarmyCV does for you: it sets up the process in minutes, collects the references without you chasing anyone, and generates the report for your client.
None of that would matter if what sat underneath were a polite “all good”. This is why you can trust the result: stars that force prioritising and justifying, open questions that test the candidate against the role, and identified voices weighted by how close they were. Rigour isn’t bolted on at the end — it’s in how every rating is collected from the very first moment.
More soon.
