1 · Read the colours first

Every KPI card and status dot uses one traffic-light system. Learn it once and you can read any page at a glance:

GREEN — on target. The metric is at or better than its goal.
AMBER — watch. Slipping toward a problem; keep an eye on it.
RED — act. Below target; it needs an owner and an action today.
GREY — informational, no target set.

Each KPI shows its value, its target (e.g. "target ≥ 90%"), a trend arrow versus the previous period, and a mini-sparkline. Click any KPI for its full definition and what to do when it turns red.

2 · The rules behind the scores (NBA v3.0)

Agents are scored by The Chatterbox NBA Scoring Framework. You do not calculate anything — the system does — but knowing the rules tells you why an agent is flagged.

The Two-Check Sales Rule

Value is always checked before Volume. Check 1 (Value): GWP per sale ≥ R99. Check 2 (Volume): daily sales vs the agent's tenure target. A high sales count at low value still fails — because it costs the business on every policy.

Value	Volume	Status → coach on
PASS	PASS	On Target → stretch goals
FAIL	PASS	Value Issue → product mix & upselling
PASS	FAIL	Volume Issue → activity, pipeline, closing
FAIL	FAIL	Both → value coaching takes priority

Tenure daily-sales targets: M1 = 3, M2 = 4, EXP-Courtesy = 5, EXP-Cuddle = 8. GWP/sale target is R99 for everyone. Agents are only ever compared with peers at the same tenure.

The Decision Ladder — one action per agent

Rules are checked top-to-bottom; the first one that triggers is the agent's single Next Best Action. Most serious always wins.

Level 1 · Protect the Business — R1 Zero Attendance · R2 Critical QA (<80%) · R2A-3 Critical Compliance · R3 Attendance Risk (<80%).
Level 2 · Protect the Revenue — R2A-2 Compliance Risk · R4 GWP Value Failure (<R99) · R4A Critical Sales-Volume · R5 Moderate QA (80–89%) · R6 Low Sales (70–89%).
Level 3 · Improve the Skill — R7 Conversion below target · R8 Talk Time (<150 min/day) · R2A-1 Compliance Note (logged only).
Level 4 · Develop & Recognise — R9 Developing (90–99%) · R10 Strong/Elite (meets all targets).

Graduated Critical-Fail (R2A)

Critical Fail Rate (CFR) = critical fails ÷ audited calls. The tier is the higher of the count band and the rate band, and the penalty is capped at −10:

Tier	Trigger	Composite	Elite
R2A-3	≥4 fails OR CFR ≥30%	−10	Blocked
R2A-2	2–3 fails OR CFR 10–<30%	−5	Blocked
R2A-1	exactly 1 fail AND CFR <10%	−2	Retained (a note, never a primary NBA)

The Composite Score (/100)

Seven weighted metrics sum to 100, then the capped critical-fail penalty is applied and reported as 'Adj': Avg GWP/Sale 25, Target Achievement 20, QA 20, Total GWP percentile 15, Conversion 10, Talk Time 5, Attendance 5. Cancellations and attempts-per-lead are diagnostics — they are never scored.

3 · How to use the software

Check Data Trust first. The bar at the top of the page shows every source's freshness — green is fresh, amber/red is stale. A beautiful chart on stale data is a confident wrong answer, so always read the trust bar before the numbers.
Upload your exports in the Upload Station. The type is detected from the headers; rejected rows are logged with reasons, never silently dropped.
Filter and set the date range in the top bar — every KPI, chart and table re-window to your selection.
Click any agent's name (on any table or chart) to open their full card — framework score, the primary NBA, the composite breakdown, graphs, peer comparison, and an AI coaching analysis of what to improve.
Ask AgentOS — the chat bubble in the bottom-right answers questions about the numbers on screen, the framework, and the business. It never invents figures.

4 · The weekly rhythm

Daily the system assigns one NBA per agent. Every Monday the Team Leader confirms the data and sets owners and due dates. During the week the coaching happens. At the next review — clear the NBA if resolved, escalate if not. If a red number has no owned action, the system has failed.