The Collab Points lens shows the explicit recognition data — the points reviewers allocated to colleagues during the cycle. Unlike the Collaboration and Culture lenses (which are AI-extracted from review text), this page works with the literal allocations.
What this page answers
Who’s the most recognised in the firm? Is recognition aligned with performance? Which departments are receiving more recognition than they’re giving? What are people being recognised for?
Before the page works
This lens requires your review cycle to include a collaboration points field. If your cycle doesn’t have one, the page will show an alert explaining what’s needed — there’s nothing to analyse without allocation data.
What’s on the page
Four headline KPIs
Departments Recognised: count of departments that received points vs total active departments
Median Points / Dept: median total received across departments
People Recognised: count of unique people who received points vs total in the cycle
Median Points / Person: median received per individual
Individual leaderboard
A sortable, searchable data table of every person who received points (10 per page by default):
Name
Department
Received: total points received
People From: count of distinct people who nominated them (breadth of recognition)
Depts From: count of distinct departments that recognised them (cross-functional visibility)
AI Performance: Compass AI performance level from extraction, colour-coded (Exceptional through Critical)
Recognition categories
A donut chart showing why people allocated points — the reason category selected when awarding. Shows what the organisation values most in peer recognition.
Points distribution
A histogram showing how many people fall into each points-received bucket (1-2, 3-5, 6-10, etc). A healthy distribution is spread across buckets. If most people cluster in a low bucket with a few outliers at the top, recognition may be concentrated on a small group.
Avg points received by performance level
A bar chart showing average collaboration points received, grouped by AI-assessed performance level (Critical through Exceptional), with the count of people per level. Shows whether peer recognition is performance-aligned — a clear upward gradient suggests recognition follows performance; a flat or inverted pattern suggests it’s driven by visibility or popularity instead.
Recognition themes
A word cloud generated from language analysis of collaboration point comments. Larger words appear more frequently. Requires language processing to have been run on this cycle’s collaboration points.
Points summary by segment
A tabbed table with three tabs (Department / Role / Location). For each segment:
People: members in that segment with a review in the cycle
Points Received: total received
People Received: how many in the segment got any points
Points Given: total given out
External Recognition: how much of the received points came from outside the segment (above 1.0 = more external than internal)
Willingness: how much of the given points went outside the segment (above 1.0 = outward-facing generosity)
Points given vs received by department
A scatter chart with X = points given, Y = points received. Departments on the diagonal give about as much as they get; departments above the line are net recipients; departments below are net givers.
Key collaboration pairs
A table of the top 10 department pairs by points exchanged. Columns: pair name, total points, people involved, reciprocity index (how balanced the exchange is — near 1.0 means equal flow in both directions).
Points flow network
A dependency wheel showing how collaboration points flow between departments. Thicker connections indicate more points flowing. The direction shows who is giving to whom.
Where the data comes from
Most values on this page come from your firm’s collaboration points allocation data. The AI Performance column in the leaderboard and the recognition vs performance chart cross-reference against Compass AI extraction data to show whether recognition aligns with assessed performance. The word cloud uses key-phrase extraction from point allocation comments.
How to read it
Start with the recognition vs performance chart. If higher performers receive more points, your recognition culture is healthy. If it’s flat, recognition may be about visibility rather than contribution.
Check the leaderboard for mismatches. Sort by AI Performance — are your top-recognised people actually your strongest performers? Look for “Exceptional” with few points (hidden contributors) or “Concerning” with many points (popularity without performance).
The scatter chart shows departmental balance. Departments far above the diagonal are recognised disproportionately to what they give back. Departments below are generous recognisers of others.
Reciprocity index in the pairs table is the most useful single number. A pair with a high reciprocity index has a healthy mutual relationship; a one-way relationship has a low one.
The dependency wheel is best for executive presentations. The leaderboard tells you who, the wheel shows leadership the shape of how the firm works together.
Tips
Sort the leaderboard by AI Performance. Quickly spot people whose recognition doesn’t match their assessed performance level.
The Department / Role / Location tabs are most useful for cross-cuts. A function that’s high on the Role tab but low on the Department tab tells you something about how the function is distributed.
Recognition data complements the AI views. Where this page says someone received a lot of points, the Collaboration or Culture lenses can tell you what they were recognised for.
Points distribution tells you about concentration. If most people are in the 1-2 bucket, recognition is highly concentrated on a few. A wider spread is healthier.