What it is
The partner scoring system is an automated monthly evaluation that measures how well collection partners perform across key dimensions. The system calculates a Quality Score from 0 to 100 by combining five performance components, then applies engagement multipliers to produce a Final Score.
The five components and their weights are:
Component | Weight | What it measures |
Ratings | 35% | Star ratings after case closure, from clients and, on managed cases, from the managing partner |
Collection | 25% | Recovery rate compared to a 25% target |
Task deadlines (SLA) | 15% | Task completion timeliness against deadlines |
Compliance | 15% | Impact of upheld complaints on partner conduct |
Support | 10% | Thumbs up/down feedback from chat interactions |
Scores are formally issued on the 1st of each month covering the previous month's performance, using all historical data with recent performance weighted more heavily than older results. Your score is also recalculated every Monday, and straight away when a complaint against you is upheld. The Monday recalculation does not trigger a partner email.
Why it matters
The scoring system exists to ensure quality and accountability across the partner network. It provides a standardized way to evaluate partner performance and helps maintain service quality for clients using the platform.
Your score reflects how effectively you recover debts, how satisfied clients are with your work, and how reliably you meet deadlines and respond to platform requirements. Maintaining a strong score demonstrates your commitment to quality collection services.
Quality Score calculation
Each component produces a score from 0 to 100. The system then calculates your Quality Score using the weighted formula:
Quality Score = (Collection x 25%) + (Ratings x 35%) + (Support x 10%) + (Task deadlines x 15%) + (Compliance x 15%)
Component details
Collection Score (25%) - Compares your recovery rate against a 25% target. Recovery is calculated as the amount collected divided by the principal claim amount on eligible closed cases, with the close-code exclusions described in Case close codes. Recovering 25% or more of principal earns a score of 100.
Ratings Score (35%) - Based on ratings collected after case closure from clients and, on a case managed by another collection partner, from that managing partner. Both rate the outcome and your communication on a 1 to 5 star scale, plus a private recommendation score (0 to 10), and a managing partner's review counts exactly like a client's. Higher ratings lead to higher scores.
Support Score (10%) - Measures the percentage of positive (thumbs up) feedback from chat interactions with clients. A higher thumbs up rate produces a higher score.
Task deadlines score (15%) - Also known as the SLA score. Tracks how quickly you complete tasks relative to their deadlines. On-time completion earns full points; being 7 or more days late results in zero points for that task. Different task types have different weights: case verification tasks count most heavily (60%), followed by chat replies (25%) and survey responses (15%). See Tasks for more on meeting your deadlines. Note: Tasks that are automatically closed by the system, either because the AI determined no reply was needed on a chat reply task or because the case was reassigned to another partner, always score as on-time and do not count against your Task deadlines score.
Compliance Score (15%) - Starts at 100 and decreases based on upheld complaints from the past 12 months. Deductions depend on severity: minor issues deduct 5 points, medium issues deduct 15 points, and severe issues deduct 40 points. Dismissed complaints and upheld complaints with No Penalty severity have no impact on the score. For complaint handling, see handling disputes with clients.
Engagement multipliers
Beyond the Quality Score, the system applies an engagement multiplier that enforces participation requirements. The Case Engagement Multiplier tracks whether you accept or ignore assigned cases.
The multiplier is calculated as: Approved cases / (Approved + Ignored cases) over the past 90 days.
If you approve all assigned cases, your multiplier is 1.0 (no penalty). If you ignore cases without declining them, the multiplier reduces your final score proportionally. Legitimate declines are excluded from the engagement calculation, so declining a case does not penalize you.
The Final Score is: Quality Score x Case Engagement Multiplier
Aging and recency
The system applies aging weights to prioritize recent performance:
Data age | Weight multiplier |
0 to 90 days | 1.5x (most recent) |
91 to 180 days | 1.2x |
181 to 365 days | 1.0x (baseline) |
366 to 730 days | 0.5x |
Over 730 days | 0.2x (historical) |
This means your recent work counts more toward your score, allowing you to improve your standing by performing well now.
Baseline scores for new partners
New partners start with baseline scores rather than zeros. The system uses a Bayesian approach that blends real data with virtual baseline data, so your score transitions smoothly as you build a track record. As you accumulate more cases, ratings, and task completions, the baseline fades and your real performance dominates.
Starting baselines by component:
Collection: 100 (assumes target recovery rate)
Ratings: 70 (equivalent to 3.5 out of 5 stars)
Support: 75 (assumes 75% positive feedback)
Task deadlines: 80 (assumes 80% on-time completion)
Compliance: 100 (innocent until a complaint is upheld)
Until you have real data, your total is a provisional blend of these baselines (84 with the default weights). The Ratings card shows a dash instead of the baseline before your first rating.
Score thresholds and colors
The platform uses a three-tier color system to indicate performance levels:
Score range | Color | Meaning |
70 to 100 | Green | Meeting or exceeding expectations |
50 to 69 | Black | Acceptable, with room for improvement |
0 to 49 | Red | Below threshold - needs attention |
This three-color scale is used on the metric report cards. On the Overview page of Partner Score, the overall score's gauge and progress use two colors instead: green at 50 or above, red below 50. This is a display rule only - it does not change how the score is calculated, and a green overall score does not mean every component is strong.
Confidence levels
Each score component includes a confidence level based on available data:
High - Substantial data available; score is reliable
Medium - Moderate data; score is becoming reliable
Low - Limited data; score may shift as more data accumulates
When your real data count is still lower than the baseline virtual count, confidence displays as Low even if you have some data. This indicates your score is still partly baseline-driven.
On the Overview, only Low is shown: the card gets a Low confidence badge, and High and Medium show no badge.
Recovery rate target
The Collection Score uses a 25% recovery rate as the target for a perfect score. This means recovering 25% or more of the principal claim amount on average earns you 100 points for that component.
Task deadline expectations
Tasks should be completed by their deadline. On-time completion earns full points. Being late reduces your Task deadlines score gradually, and being 7 or more days late results in zero points for that task.
Tasks closed automatically by the platform, including chat reply tasks auto-resolved by AI when no reply was needed and tasks closed when a case is reassigned to another partner, always receive a full score and are not penalized for lateness.
Where to find your scores
View your performance metrics on the Partner Score page in the Partner Portal. It opens on Overview, which shows your overall score with its gauge and the date it was last calculated, a How your Partner Score works explanation you can expand, and five category cards. While your overall score is still based on limited data, it carries a Provisional badge with "Based on limited data". If ignored cases reduce your score, a banner shows the figure before and after, for example "Your score is reduced from 61 to 53 because you have ignored cases."
Each card shows the component score and ends with a link to its report or, for Task deadlines, to Tasks. If you have overdue tasks, the Task deadlines card shows how many and links straight to them. A card shows a Low confidence badge only when that component is based on limited data; cards with enough data show no badge.
The report pages under Partner Score are:
Collection Report - Collection Score, confidence level, recovery rate, and average days to payment
Received ratings - Ratings Score, confidence level, and average outcome and communication stars, with each rating listed by case and reviewer
Complaints - Compliance Score, complaints with impact, response rate, and open complaints
Support Feedback - Chat Feedback Score, confidence level, and feedback breakdown
Each report page highlights the main score in a featured card, with supporting metrics providing additional context.