Skip to content
HeuriTel
Sign in

Mobile financial services, bank and MFI credit risk teams

Estimate who will repay, and let the lender decide

A wallet customer asks for a small loan before payday. The lender holds an approved, purpose-limited view of their operator and wallet history, and has to answer in seconds.

Read the journey; a working session shows it running.

Seen by the customer · USSD

From the record, not a run: what would reach them, on USSD, App, Agent.

The case

Annotated from the record

What happened

A wallet customer asks for a small loan before payday, and the lender has an approved, purpose-limited view of their operator and wallet history.

A wallet customer asks for a small loan before payday. The lender holds an approved, purpose-limited view of their operator and wallet history, and has to answer in seconds.

What it reads

  • Consent and the approved, purpose-limited attributes
  • Wallet and operator history within the approved window
  • Affordability and existing obligations
  • The lender's policy and limits

Checks before anything is sent

Consent, purpose limitation, data approval, affordability and the lender's own policy checks pass.

Every option considered, including doing nothing

  1. Approve at the recommended limitConsidered
  2. Approve at a lower limit, with the reasonConsidered
  3. Refer for a human decisionConsidered
  4. No loan, with the reason codes shownConsidered

Doing nothing is one of the outcomes, and it is recorded with its reason. The options are the record's own; a run decides between them.

Evidence record

Preview
Reference
The reference of the saved run.
Decision
One of the options, including no action, with the reason.
Checks
Consent, purpose limitation, data approval, affordability and the lender's own policy checks pass.
Evidence kept
  • Consent
  • approved attributes
  • estimate
  • reason codes
  • limit recommendation
  • lender's decision
  • disbursement
  • repayment record
The measures
Repayment on time against the propensity estimated, and delinquency caught early.
Connection
Written on every record: the systems the run read, and the sample it ran on.

The shape of the record, with no values. A reference appears only after a run is saved.

The improvement, and how it is measured

Hypothesis

A loan sized to repayment is repaid and renewed; one sized to the request is a collection.

Modelled

The mechanism that could produce the result, stated; the figure is measured in the pilot.

Observed

Measured in a pilot’s validate stage, on your systems, against a comparison agreed first.

The result, the mechanism, the measurement, and the record’s five answers
Desired result
Incremental revenue or retention; lower contact waste; improved conversion with margin guardrails
Mechanism
Ingest → qualify → score → apply limits/consent → assign control → execute approved action → capture delivery and business outcome → monitor/retrain
Measurement

Randomised holdout where feasible; intent-to-treat primary view; pre-declared denominator; guardrails for complaints, margin and opt-out

Measured on: Eligible population; treatment rate; conversion; incremental revenue; ARPU; churn; contact rate; margin; opt-out; decision latency

  1. 01
    Desired result

    The result wanted.

    Incremental revenue or retention; lower contact waste; improved conversion with margin guardrails

  2. 02
    Mechanism

    The mechanism that could produce it.

    Ingest → qualify → score → apply limits/consent → assign control → execute approved action → capture delivery and business outcome → monitor/retrain

  3. 03
    Evidence required

    The records kept to show it.

    Decision log; eligibility snapshot; price/order response; delivery receipt; control assignment; revenue/outcome ledger

  4. 04
    Costs and risks

    The cost, and the ways it can go wrong.

    Costs: Setup fee + annual product subscription; optional managed execution fee; optional verified-outcome fee with agreed baseline

    If it fails: Fail closed on eligibility/consent/price; queue retryable events; do not duplicate orders; route exception to campaign operations

  5. 05
    Measurement approach

    The measure that shows it helped.

    Randomised holdout where feasible; intent-to-treat primary view; pre-declared denominator; guardrails for complaints, margin and opt-out

    Measured on: Eligible population; treatment rate; conversion; incremental revenue; ARPU; churn; contact rate; margin; opt-out; decision latency

Measured in the pilot's validate stage, on your systems, with the comparison agreed first. Every line above is the record's own design intent.

The business side

The change
Repayment on time against the propensity estimated, and delinquency caught early.
Running cost
Setup fee + annual product subscription; optional managed execution fee; optional verified-outcome fee with agreed baseline
What evidence supports it
Consent, approved attributes, estimate and reason codes, limit recommendation, lender's decision, disbursement and repayment record.

The journey in detail

Step by step, from both sides
StepWhat the customer experiencesWhat the operator does
The momentThe moment that started it.A wallet customer asks for a small loan before payday, and the lender has an approved, purpose-limited view of their operator and wallet history.Reads consent and the approved, purpose-limited attributes, wallet and operator history within the approved window, affordability and existing obligations, the lender's policy and limits.
The decisionThe decision to be made.The customer's screen is unchanged.Estimates the calibrated probability of repayment from the approved attributes, with reason codes, and recommends a limit; the lender decides.
The safeguardsThe conditions that stop it.Still unchanged: the action waits until every check has passed.Consent, purpose limitation, data approval, affordability and the lender's own policy checks pass.
The actionThe action that reaches the customer.The lender's credit engine approves, limits or declines; the customer sees the decision and its reasons in their own language. Reaches them on USSD, App, Agent.Recommends. The system that holds the right confirms, charges or provisions.
When it goes wrongThe failure, and the recovery.The estimate reads an attribute the purpose approval did not cover, and a decision is made on data the customer did not consent to.The purpose-limitation check refuses the attribute before the estimate runs, the decision is made on the approved set, and the audit records which attribute was excluded and why.
The resultThe change it made.What changed for them is what is counted; nothing else is claimed.Repayment on time against the propensity estimated, and delinquency caught early.
The proofThe proof anyone can check.Can be answered for, later, from the record.Consent, approved attributes, estimate and reason codes, limit recommendation, lender's decision, disbursement and repayment record.

Operator systems remain authoritative. HeuriTel reads these to propose an action; it does not decide on their behalf.

The failure, and the recovery

What goes wrong

The estimate reads an attribute the purpose approval did not cover, and a decision is made on data the customer did not consent to.

What happens then

The purpose-limitation check refuses the attribute before the estimate runs, the decision is made on the approved set, and the audit records which attribute was excluded and why.

The products in this journey

Available to see today

Read the journey; a working session shows it running.

See HeuriTel on the sample of your market

Verify your email to open a demonstration on the sample of your market.