Pricing, propensity and offers.
What should this customer be offered, and at what price? Luminnovate builds each price from that customer's own expected loss and cost to serve, and ranks offers within the rules you set.
Prices built from a population average overcharge some customers and undercharge others. A pre approved offer is an invitation, not a decision, and has to be explainable to the customer who receives it.
How the rate is built
| Cost of funds | 6.10% |
| Expected loss, this customer | 2.40% |
| Cost to serve, this customer | 0.90% |
| Capital charge | 1.20% |
| Margin | 2.50% |
| Rate | 13.10% |
Offers ranked for this customer
| Line of credit, $60k | best fit |
| Term loan, $150k | refer |
| Equipment finance | not offered |
Each offer carries its reason and the evidence behind it.
Illustrative example. Harbour Joinery is a synthetic business.
What Luminnovate does
- Builds the price for one customerCost of funds, expected loss, cost to serve, capital and margin, each shown.
- Ranks offers within your rulesThe objective you choose, within limits on risk, exposure and fairness.
- Explains every offerEach offer carries its reasons and the evidence behind them.
Three questions behind every offer
Each is its own model, and they are asked in this order, with eligibility and risk first.
- Can they afford it?Eligibility and risk come first. Customers showing hardship indicators are designed to be excluded from marketing altogether.
- Who is likely to accept?Propensity: the chance a customer takes up the offer.
- Who accepts because of the offer?Uplift: customers who would accept only because they were asked, not those who would have come anyway.
A random holdout measures what the offers actually changed, the same way the Prove value capability does. Offers are invitations with reasons, never automatic approvals.
Start with one product you already price. We rebuild its prices on your own history and show where they would differ.
Book a conversation