Predict who will pay.
A 0–100 probability score computed from real engagement data. Know which debtors are likely to settle — and which need escalation.
- Real-time probability scoring
- Payment history weighting
- Behavioural pattern analysis
- AI-driven strategy adaptation
Hartley Brooks Ltd
10%Nova Services
30%Meridian Corp
55%Vertex Holdings
40%Atlas Group
85%Four probability bands
Each band maps to a distinct recovery strategy.
Very Likely
80–100%
Payment received or strong engagement history. Debtor is actively settling.
- At least one payment received
- Consistent engagement with payment pages
- History of responsive behaviour
Probable
50–79%
Debtor is clicking links and visiting payment pages but hasn't paid yet.
- Link clicks detected
- Payment page visits recorded
- Active portal engagement
Uncertain
30–49%
Some email opens but no meaningful engagement. Debtor is aware but not acting.
- Emails opened but no clicks
- No payment page visits
- Passive awareness only
Unlikely
0–29%
Ghosting or completely unresponsive. Multiple attempts with zero engagement.
- Zero opens or reads
- 3+ messages ignored
- Ghosting state confirmed
What drives the score
Six weighted factors that compute every debtor's payment probability.
Payment history
Any payment received jumps the score to 80+. Partial payments still count as a strong positive signal.
Click behaviour
Link clicks and payment page visits indicate intent. Scores jump to 55+ when clicks are detected.
Open & read activity
Email opens and WhatsApp reads show awareness. Score reaches 40 with opens but no further action.
Ghosting pattern
Confirmed ghosting drops the score to 10. Complete silence is the strongest negative signal.
Communication volume
More messages sent without response lowers the score. The system measures attempts vs engagement ratio.
Behavioural profile
Tags like 'quick_opener' boost the score. Tags like 'unresponsive' or 'opens_but_ignores' reduce it.
Probability scores across the Recovery Engine
Payment probability drives strategy at every stage.