The Overview page is the first stop on the Reviews tab. It’s designed to be an executive briefing: in ten seconds you should know how the cohort is performing, where the strengths and growth areas are, and whether anything looks unusual.
What this page answers
How is the cohort distributed across performance levels? What’s the average sentiment of the reviews? Where do self and manager views diverge? What themes dominate?
Pick a cycle first
The Overview (and every Reviews page) only makes sense for a specific review cycle. Use the cycle picker in the sidebar to select one. If you don’t, the page will prompt you to.
What’s on the page
Three headline KPIs (top row)
Reviews Extracted: how many of the cycle’s completed reviews Compass has analysed, vs the total completed. If this is less than 100%, the queue is still processing.
Avg Review Quality: the mean of Compass’s 0–1 overall quality score across all extracted reviews.
Avg Sentiment: the mean of Compass’s −1 to +1 sentiment score across all extracted reviews. Positive means reviews trend supportive; negative means more critical.
How is the cohort performing?
A bar chart showing the count of employees in each performance level: Exceptional, Strong, Solid, Concerning, Critical, Insufficient Data. This is Compass’s read on each individual review — a cohort-shape view of performance.
Sentiment distribution by department
A jittered scatter plot. Each dot is a person, grouped by department, coloured by performance level. Vertical position shows sentiment; spread shows variance. Click a dot to open that person’s detail popup.
Useful for spotting departments where sentiment is uniformly low (a culture or manager issue) versus departments where it’s bimodal (some loved, some not).
Perspective alignment
A donut chart of the self-awareness gap distribution across the cohort:
Aligned: self and manager agree
Over-rater: self rates higher than manager
Under-rater: self rates lower than manager
Mixed: inconsistent
Insufficient data
Review quality by department
A grouped bar chart with two series per department: Overall Quality (the quality of the review content) and Reviewer Quality (how well the reviewer wrote the review). Comparing the two helps spot whether the issue is content or writing.
Sentiment drivers
Bar chart of the average sentiment when each theme is discussed. Themes with very negative average sentiment are the topics dragging the cohort down; very positive themes are the ones lifting it.
Key contributions
A lollipop chart of the top 15 contribution descriptors — the words and phrases Compass found most often when describing what people are doing.
Top strengths
A lollipop chart of the most common strengths identified across the cohort (top 15).
Top development areas
A lollipop chart of the most common primary development needs (top 15). This is the cohort-level read on where coaching investment will pay off.
Where the data comes from
Every signal on this page is produced by the review extraction job — Compass reads each completed review’s content (markdown, including ratings and free-text answers) and emits the structured signals. Aggregation is done live as you load the page; the underlying analyses are cached per review.
If a cycle’s extraction coverage is below 100%, the Introduction page shows a queue status banner with the count of extraction jobs still processing or queued.
How to read it
Start with the performance distribution. Is the cohort weighted to Strong/Solid (healthy) or piling up in Concerning/Critical (a problem)?
Read Avg Sentiment alongside it. A negative average with a healthy performance distribution might mean reviewers are being unusually tough; a positive average with poor performance might mean reviewers are sugar-coating.
Use perspective alignment to gauge calibration. A cohort where most people are “aligned” is well-calibrated; a cohort dominated by “over-rater” suggests employees are evaluating themselves more generously than their managers.
Top development areas tell you where to invest. If the same gaps come up across hundreds of reviews, that’s a firm-wide training need, not an individual issue.
Tips
Compare to last cycle. The numbers on this page only tell a full story when compared to a prior cycle. The Compare Reviews tab does this side-by-side.
Filter by department to compare segments. A bad sentiment in one department but healthy elsewhere is much more actionable than a flat cohort view.
Drill in from here. Top strengths and development areas are the headline; the Reviews > Development page goes deeper.