Ask what “validated” means, and most hiring platforms give you a shrug dressed up as confidence.
Validation isn't a badge.
It’s a process and here’s where it stands.
We’re not claiming certainty. We’re raising the odds.
What we can say today: S9 InnerView measures real, job-relevant behavior and scores it consistently.
What we can’t say yet: that a high score guarantees a great hire. That takes real people, in real jobs, over real time and that’s the data we’re collecting right now.
What does “validated” actually mean here?
People use the word “validated” to mean at least three different things, and we’re not always careful about which one we mean. Are the right skills being measured? Does the same evidence produce the same score twice? Does a good score actually predict a good hire? Those are three separate claims, at three different stages of being answered, and this page keeps them separate rather than folding them into one reassuring sentence.
Content validity
Are the right skills being measured for this role?
Actively reviewed against real role requirements from design partners.
Scoring consistency
Does similar evidence get scored similarly across interviews?
Actively measured during the beta, as more interviews happen.
Predictive validity
Do Stratum Scores actually predict on-the-job performance?
Not yet established. This is specifically what “in progress” refers to.
What’s being measured right now, during the beta?
During this design-partner phase, we’re actively checking two things.
First, whether each of the 45 skills genuinely belongs to the role it’s assigned to a Culture skill picked for a sales role should actually matter for that job, not just sound relevant.
Second, whether the same kind of evidence gets scored the same way, regardless of which candidate, which interviewer, or which interview it came from.
Getting both right is groundwork. It isn’t proof that a higher score leads to a better hire that’s a separate, harder question.
Why hasn’t predictive validity been established yet?
Because it can’t be rushed. Answering it requires people who took the job, stayed in it, and did the work long enough that their performance can actually be judged. A beta running for a few months hasn’t produced that yet. As design partners hire through the platform, those hires accumulate real track records and that takes time passing in the real world, not a faster query.
What happens as more data comes in?
It gets published. Including the version where the results aren’t flattering.
A validation page that only ever reports good news isn’t validation.
It’s marketing wearing a lab coat.
What’s in the Science packet?
More detail than this page has room for: the full methodology, what’s specifically being tracked right now, and just as importantly a plain list of what we’re not claiming yet. If you have to walk this into a conversation with your own legal or compliance team, this is what you’d actually bring.
Frequently asked questions
No, not yet. That question specifically needs a group of hires who’ve been in their roles long enough to evaluate, and we’re in the process of collecting that now. We won’t call it proven until the data actually says so.
Right now, two things: whether each skill actually belongs to the role it’s applied to, and whether similar evidence gets scored consistently across different interviews. Whether the scores predict performance comes after that, once there’s enough real-world outcome data.
There’s no date we can give you honestly. It depends on how quickly design partners hire and how long it takes before someone’s performance can fairly be judged. We’d rather wait than hand you a number we made up.
Everything above, in more detail: the methodology, what’s actually being tracked, and where we’ve drawn the line on what we’re willing to claim.
Still here? Respect.
You've read more than most interviewers read job descriptions.
Take five, here's something to listen to that isn't a cue card.