Free · no email required
Six learners. One paragraph. Six different readings.
The panel is easy to describe and hard to believe until you see it work. So we gave all six profiles the same 74 words from a course we actually audited.
The specimen
One paragraph, taken verbatim.
This is a single instructional page from a mandatory compliance course, Case A in our published runs. It is the second thing a learner reads. Nothing has been edited except the name of the statute.
A human reviewer reads it and thinks: accurate, complete, a bit dry. It is accurate. It is complete. It is also lifted straight from the statute. Five profiles find five different problems in it. One finds nothing wrong. That is useful too.
The panel
What each one said.
The paragraph stays where it is. What is marked on it changes as you read down.
Case A · module 1, page 2 of 23
Benefit of Employment. [The statute] defines “benefit of employment” as: any advantage, profit, privilege, gain, status, account, or interest (other than wages or salary for work performed) that accrues by reason of an employment contract or agreement, or employer policy, plan, or practice, and includes rights and benefits under a pension plan, health plan, or employee stock ownership plan, as well as insurance coverage and awards, bonuses, severance pay, supplemental unemployment benefits, vacations, and the opportunity to select work hours or location of employment.
One sentence · 74 words · 17 enumerated items · no example · nothing on the page is assessed
Cold StartSeventeen things it is, one thing it isn't, and no way to reason about anything else.
Cold StartNo prior knowledgeInstructional“I can read every word of this and still not know what isn't a benefit of employment.”
A definition that enumerates seventeen things and gives one exclusion teaches recognition of the list, not the concept. The learner’s actual task is to decide whether something not on the list qualifies, and the page provides nothing to reason with.
The PractitionerFifteen years in the roleInstructional“It lists bonuses. It doesn't say whether a discretionary bonus counts, and that is the only reason I would ever open this page.”
Experienced learners do not come for the definition. They come for the edge that made them uncertain. A page that restates the rule without addressing where the rule is hard is a page they will stop trusting, and then stop reading.
Cold Start
I don’t understand where the concept ends.
The Practitioner
I understand the concept. Tell me where it gets difficult.
Same paragraph, opposite instructional needs, and a fix for one is not a fix for the other.
The Mobile LearnerTouch only, small screenAccess“Seventy-four words as one unbroken sentence, on a page with a fixed 796-pixel container and no viewport tag.”
On a phone the fixed-width page scales down rather than reflowing, which makes an already dense sentence harder to read and harder to navigate. The LMS can record whether someone progressed through the course. It cannot tell you the page was painful to use.
The Second-Language ReaderFluent, but not nativeAccess“'Accrues by reason of'. I know all three words and I have to stop and rebuild the clause.”
One sentence, seventeen coordinated items, a parenthetical exclusion in the middle, and a register two levels above the task. Comprehension does not fail outright; it degrades quietly, and nothing downstream ever checks whether it did.
The Keyboard-Only UserNo mouse, no touchAccessNo finding“Nothing wrong here. Skip link present, next control reachable, focus visible.”
This is a real result and we report it as one. Not every profile should find a problem. If they always do, the panel is telling us what we expect instead of what is actually there. On this page, this learner passes cleanly.
Minimum EffortWants to pass and leaveHard gate“Nothing on this page is assessed. Next.”
The most consequential result, and it takes four words. The other five profiles ask whether the learner can learn from the instruction. This one asks whether the learner needs the instruction at all. It does not evaluate the writing; it evaluates whether the writing can be skipped without penalty. On Case A the answer was yes throughout the course, and the course failed outright as a result.
It is worth being plain about why this profile exists, because it is the one people find uncomfortable. It is not a portrait of a bad learner. It is what a reasonable person becomes after the fourth course that could have been an email. A course that cannot survive it was relying on goodwill rather than on design, and the question it really asks is whether this course needed to exist at all.
The point
Six readings of the same 74 words.
The paragraph hasn’t changed. The question being asked of it has.
- Cold Start asks whether a novice has enough context.
- The Practitioner asks whether an expert gets anything useful.
- The Mobile Learner asks whether the instruction survives the device.
- The Second-Language Reader asks what the language itself demands.
- The Keyboard-Only User asks whether the interaction remains operable.
- Minimum Effort asks whether any of it has to be learned at all.
That is the point of the panel: not to replace a good reviewer, but to make six different failure conditions part of every review.
A review is only as broad as the perspectives it deliberately tests. Running the same six every time is what makes the coverage a property of the process rather than of the day somebody had.
What this page is not
It is one page of one course. What you have just read is the mechanic, not the method.
This demonstration shows what the profiles surface. A full Preflight adds repeated runs across an entire course, course-wide analysis, scoring against seven dimensions, source tracing for every finding, human verification, the hard gate, ranked remediation and a retest. The scoring rubric and its cut lines are published, and so are the sources behind every check.
What would the panel find in your course?
Six profiles, three runs each, eighteen documented runs and a scored report in ten business days.