Think of the best technical salesperson you know, promoted to manager because she was brilliant with clients. She learned to run the team, of course. The results came. Her references, if you asked for them today, would be excellent.
Now ask her how she feels come Friday.
Some people do a job very well and it charges their batteries; others do a job just as well and it drains them. From the outside they’re indistinguishable: same results, same references, same performance. The difference lives somewhere no hiring process looks.
“Having or lacking a competency doesn’t stop you doing a job. It determines how much energy the job costs you.”
That’s the idea running through this whole series. And it has a direct consequence for anyone who hires: two candidates with identical references can be completely different bets.
✓ Candidate A
Leadership confirmed · ★★★★★
It’s also her natural ground
Performs — and it’s sustainable
⚠ Candidate B
Leadership confirmed · ★★★★★
Runs on sheer willpower
Performs today — at what cost tomorrow?
With the usual tools, A and B are the same candidate. Same stars, same references, same report. What sets them apart — the internal cost of that performance — was, until now, invisible.
The two signals
Seeing that difference takes two independent sources of information about the same person:
👥
The observed signal
What the people who worked with someone confirmed about their skills: structured references, with examples, evidence and verified identity. It says what they demonstrated.
🧭
The inferred signal
What the person is naturally suited for: the kind of work that costs them the least energy. It says where their natural ground lies.
The first signal is what we spent our whole previous series building: structured, verified references. It’s solid — but it looks at one side only: the demonstrated past.
The second has existed for decades in the shape of behavioural tests. On its own, though, it has a serious problem.
Why one signal alone isn’t enough
An isolated behavioural test tells you someone “tends to be organised” — and there it stays, floating, checked against nothing real. Nobody confirms whether that predisposition ever turned into actual work. That’s why so many tests end up as a friendly label nobody uses to decide.
And references alone have the mirror problem: they tell you what the person demonstrated, but not at what cost. They can’t tell candidate A from candidate B.
“One signal without the other only gets you halfway. The new information isn’t in either of them — it’s in the crossing.”
When the two signals agree, you decide with a confidence no single source could give you: what they demonstrated is also their natural ground.
When they diverge, information appears that no hiring process used to have: someone excellent at something that drains them (how long can they sustain it?), or someone clearly suited for something they haven’t yet had the chance to prove (how much potential sits there?). Divergence doesn’t rule anyone out — it informs the decision.
For the first time, in the same place
This is what we’ve just released on StarmyCV: the second signal, integrated alongside the references you already know. Each person can discover their behavioural profile, and you as a recruiter can cross both signals in your processes and in the report you hand your client.
Test providers don’t have verified references. Reference tools don’t have a behavioural signal. For the first time, both live and get contrasted in one place — and that changes what a hiring process can know about a person.
What’s coming in this series
- 1What the behavioural signal looks like inside: from a short questionnaire to a profile with concrete competencies — no lions, no colours, no labels.
- 2Behavioural fit inside your process: you define what the role needs and the system crosses each candidate — straight into your report.
- 3When the two signals don’t agree: how to read divergences, which is where the most valuable information lives.
We start next week with the first one. If you select people for a living, this series will hand you a reading tool that didn’t exist until now.
More soon.
