Skip to content
Qwackly

Risk Factor Scoring

Weigh the risk your way. Escalate what falls outside it.

Define the factors that matter to your book, how much each one counts and which score bands you are comfortable quoting automatically. Everything else is flagged for a person, with the reason written down.

Risk scoring · Q-24187
Band B
FactorWeightPoints
  • Claims history (3 yrs)3012
  • Vehicle age209
  • City risk zone2011
  • Usage type154
  • Security fitment153

Composite score

39/100

A · 0-25Auto-quote, +0%
B · 26-45Auto-quote, +9%
C · 46-65Auto-quote, +22%
D · 66-80Auto-quote, +38%
E · 81+Refer to underwriter

Capabilities

What you get out of the box

Weighted factors you control

Claims history, asset age, location zone, usage and protective measures each get a weight and a points scale. Change a weight and every future quote reflects it, with the previous version kept for audit.

Escalation, not guesswork

Scores above your threshold, or specific factor combinations, are referred to a reviewer instead of quoted.

Score bands to loadings

Map each band to a loading percentage, so the breakdown can show exactly why risk moved the price.

Reviewer queue

Referred cases arrive with the intake, the score and the triggering rule, ready for a decision.

Consistent across agents

The same answers always produce the same score, whoever is quoting and from wherever.

How it runs

From answers to a band, and a decision.

Scoring runs between intake and pricing. Its output decides both the loading and whether the quote can be issued automatically.

  1. Step 01

    Collect factor inputs

    Intake answers are normalised into factor values, such as years since last claim or city zone.

  2. Step 02

    Score each factor

    Each value is converted to points on the scale you set, then multiplied by the factor's weight.

  3. Step 03

    Place in a band

    The composite score lands in a band. Each band carries a loading and an auto-quote rule.

  4. Step 04

    Quote or refer

    Within standard rules, the quote goes out instantly. Outside them, it goes to a named reviewer.

In detail

Escalation rules, stated plainly

Referral is never a silent failure. The reviewer and the customer-facing agent both see why a case was held.

Score threshold

Any composite score above the limit you set for a line is referred.

Score 81+ on private car

Factor combination

Specific pairings that your underwriters want to see, regardless of the total.

High risk with comprehensive cover

Hard stop

Answers that make a case ineligible for standard cover under your rules.

Commercial use on a private policy

Questions

Asked by teams like yours

Does the AI decide who gets cover?
No. Scoring applies the weights and thresholds you configure. Anything outside your standard rules is referred to a human reviewer, and the final decision on those cases stays with your underwriters.
Can we use different factors for different lines?
Yes. Each line of business has its own factor set, weights and bands. A home quote does not need to share anything with a motor quote.
How do we test a weight change before it goes live?
Run the impact preview. It re-scores a sample of recent quotes under the new weights and shows how many would move band, and by how much the premium changes.
Can a reviewer override a referred case?
A reviewer can approve, decline or adjust a referred quote. Any adjustment is recorded as a manual line in the breakdown with the reviewer's note.

Your turn

Price one yourself. Watch it explain itself.

SAMPLE QUOTE · ILLUSTRATIVE RATES

Q-SAMPLE-01

Line of business
Coverage level
Risk profile

Loyalty

Base rate
₹4,200
Risk adjustment
+ ₹380
Loyalty discount
− ₹150
Total
₹4,430

This is a sample quote using illustrative rates. Connect your own rate rules for real quotes.