A baseline answers a question most organisations cannot: where does our workforce actually stand today? Everything the module claims afterwards — that capability improved, that training worked, that someone is ready for the next role — is measured against it.
Get the baseline right and the rest follows. Get it wrong and you spend two years comparing against a number nobody trusts.
Before you start
A cycle can only assess what the Skills Genome already describes. Confirm all four:
- A competency framework exists and is active (not draft).
- Job roles exist, with required levels set on their competencies.
- People are in the workforce register with a job role assigned.
- People have a manager set, if you want supervisor assessments.
A person with no job role cannot be enrolled — there is nothing to assess them against. A person with no manager gets a self-assessment but no supervisor one. Both are reported back to you at enrolment rather than silently skipped.
Step 1 — Create the cycle
Workforce → Assessment Cycles → New cycle.
Choose the type carefully. Baseline is special: closing a baseline cycle freezes a permanent reference point. Every future readiness percentage is compared against it, and it cannot be recalculated later even if you change the framework or the engine weights.
Scope it to an org unit if you are rolling out gradually. A pilot in one department is usually wiser than a whole-organisation baseline on day one.
Decide who rates:
- Self-assessment — the employee’s own view. Weighted lightly, but the gap between self and supervisor is often the most useful conversation in the whole exercise.
- Supervisor assessment — the substantive judgement, weighted much higher when it carries evidence.
Step 2 — Enrol people
Enrol people creates the assessment sheets. You will get back three numbers: how many people were considered, how many sheets were created, and how many people were skipped with the reason.
Read the skipped list. It is almost always telling you that your job architecture or reporting lines are incomplete, and that is much cheaper to fix now than after the baseline is frozen.
Step 3 — Open the cycle
Opening makes the sheets available to raters. A cycle cannot be opened with no assessments in it.
Supervisors see their sheets under My Team → Assess, employees under My Development → My Self-Assessment.
Step 4 — The rating itself
This is where the module differs from an ordinary performance tool. A rater does not “score someone 1 to 5”. They are shown, for each competency, what each level looks like as observable workplace behaviour, and asked which one they have actually seen.
Encourage raters to fill in the observation note. “Diagnosed the hydraulic fault on line 3 unaided on 12 August” is what makes a rating survive a disagreement six months later. “Good at diagnostics” does not.
Where formal evidence exists — an assessment result, a portfolio item, a logbook entry — attach it. A supervisor rating with evidence is weighted at 0.85; the same rating without evidence is weighted at 0.60. That difference is visible to everyone and is deliberate.
Step 5 — Watch what the numbers do
Scores recalculate the moment an assessment is submitted. Two things will surprise people on the first run, and both are correct:
Most competencies will show an em dash, not a number. A score is only published once enough verified weight has accumulated. One self-rating is not enough. This is the Suggested/Verified firewall, and it is what stops the system asserting things it cannot support.
Readiness will look lower than expected. It is capped at 79% while any mandatory competency is unmet, and unverified competencies count as zero toward it. “We don’t know” is not “ready”.
Both resolve as evidence accumulates. Resist the urge to lower your standards to make the first numbers look better.
Step 6 — Calibrate before you close
Once supervisor assessments are in, run a Calibration session. It puts each manager’s ratings next to the evidence-based score and the peer median, and shows you which managers are consistently more generous or more severe than their peers.
The system surfaces the pattern. A human decides what it means. An adjustment requires a written reason and is recorded as a new rating — the manager’s original judgement is kept, because the fact that they made it is itself part of the record.
Step 7 — Close the cycle
Closing freezes an immutable snapshot for everyone in it. For a baseline cycle that snapshot is permanent.
Close it even if some assessments are incomplete. A partial baseline is still an honest record of what was known on that date, and a baseline that never closes is worth nothing at all.
What you have afterwards
- Capability Heatmap — where the organisation stands against what its roles require.
- My Growth — each employee’s readiness from baseline to today.
- Gap analysis — exportable to PDF, Excel or CSV, with unmeasured competencies shown as em dashes in every format.
- A challenge path — employees can acknowledge, comment, disagree, request evidence, request a reassessment, or appeal.
That last one matters more than it looks. A system that scores people without a way to contest the score is not defensible, and every rating in the module carries an attributable author and a timestamp precisely so it can be argued with.