Skip to main content
The pre-read helps participants focus on the reviews most likely to need discussion. Windy analyzes the included review packets and surfaces data-driven observations for the calibration committee. The pre-read does not make calibration decisions. Participants should open the relevant packets and confirm the evidence before changing a rating.
The pre-read is read-only. You cannot edit Windy’s report. If its inputs change, update the manager reviews or calibration settings and regenerate the report.

​
What the pre-read highlights

Windy looks for patterns such as:
  • A rating that does not appear to match the written evidence
  • Differences across manager, self, peer, or upward feedback
  • A manager whose ratings trend higher or lower than peers’
  • Inconsistency in how teams apply a rating scale
  • Missing or limited support for a rating
  • Feedback that may be vague, difficult to act on, or misaligned with development goals
  • Reviewees or questions that merit closer committee attention
When available, Windy also uses supplemental data as context. The report reflects the calibration roster, visible questions, packet contents, and supplemental data at the time it is generated.

​
Review and share the pre-read

1

Wait for the initial report

Windmill starts generating the pre-read automatically when the calibration is created. Generation happens in the background and may take a few minutes.
2

Open the pre-read

Open the calibration, then select Pre-read in the right rail.
3

Review or refresh it

Review the report before distributing it. If generation failed or relevant inputs changed, a facilitator can generate a fresh version.
4

Share it with participants

To email the report, select Send. This emails every facilitator and committee member in the calibration. The send flow shows the recipients and lets you include an optional message. You can also copy the report or download it as a PDF.
Facilitators decide when to send the pre-read. Review it late enough to include most manager submissions while still giving participants time to prepare.

​
Regenerate an outdated pre-read

The pre-read may become outdated when relevant inputs change, such as:
  • A manager submits or updates a review
  • A facilitator changes which questions are visible
  • The reviewee roster changes
  • Supplemental data or its fields change
When Windmill marks the pre-read as outdated, generate a fresh version before participants rely on it. The new version replaces the report attached to the calibration.

​
Run an effective pre-read process

  • Send the rating rubric with the report.
  • Ask participants to review the highest-priority cases before the session.
  • Have participants add packet comments for questions that require context.
  • Separate data-quality issues from performance decisions.
  • Regenerate after material late submissions.
Example: Windy flags that one manager’s team has unusually high ratings and that two ratings have limited written support. Participants review those packets asynchronously, leave questions for the managers, and reserve live meeting time for the cases where the evidence remains unclear.

​
Access and privacy

Only facilitators and committee members in the calibration can view its pre-read. Reviewees and cycle admins who are not participants cannot access it. The pre-read can contain sensitive performance information. Send it only through the calibration’s participant-sharing flow and avoid copying it into broader channels.

​
FAQs

Yes. Windmill requests the initial pre-read automatically when the calibration is created. A facilitator chooses when to refresh it and when to send it to calibration participants.
No. Selecting Send emails the pre-read to every facilitator and committee member in the calibration; reviewees never receive it. Sending is manual, so a facilitator chooses whether and when to send it.
No. Committee members can view the report, but only facilitators can generate or regenerate it.
No. The pre-read is a read-only report generated from the calibration’s current inputs. Update the source reviews, roster, visible questions, or supplemental data, then generate a fresh version.
No. It identifies patterns for the committee to investigate. A facilitator must explicitly propose any rating change.
One or more inputs changed after the current report was generated. Refresh it to include the latest review data, roster, visible questions, and supplemental data.
This section evaluates the quality of the developmental feedback managers gave — not the employee’s performance. It has three parts:
  • a) Manager Feedback Actionability Score (1–5). Each manager review gets a score with a one-sentence justification, using this rubric:
    • 4–5 (High): specific, behavioral examples; links behavior to business impact; suggests concrete, forward-looking actions
    • 3 (Medium): vague suggestions without clear examples
    • 1–2 (Low): relies on clichés or focuses on unchangeable personality traits
  • b) Developmental Alignment. Flags employees where the manager’s feedback themes don’t line up with what the employee themselves identified as development areas in their self-review.
  • c) Development Gaps for High Performers. Specifically flags top-rated employees whose Actionability Score is 3 or lower — i.e., strong performers who received low-quality developmental feedback and risk being under-coached.
So the score is a 1–5 rating of how actionable and specific the written feedback is, with 5 being the max, each accompanied by a short rationale.