All use cases

Upsell & Cross-sell

Banking product cross-sell from account signals

A customer's balance has sat idle for months while a savings product goes unmentioned — the agent makes the offer at the moment the account itself makes the case.

Get aheadNext best thingBehavioral triggerSystem triggerIn-appChatEmailFintech

What it does

Most banking customers hold one product while their account activity quietly argues for a second: an idle balance that should be earning interest, recurring international transfers paying full FX fees, card spend concentrated in a category a better product rewards. Banks see these signals but act on them with batch campaigns weeks later, if at all, and the moment passes. The agent watches account behavior, product usage, and life-event signals, and opens a personal recommendation the moment a signal fires — the right product, explained against the customer's own activity rather than a segment's.

How it works

  1. 1
    Trigger. an account signal fires — a balance crossing an idle threshold, recurring foreign-currency transfers, a salary-deposit change, repeated card spend in a category another product serves better.
  2. 2
    Decision. the agent checks the customer's memory (products already held, offers previously declined, past conversations), then eligibility and suitability policy; a judge gates every unprompted message so only a genuinely fitting recommendation goes out.
  3. 3
    Action. a short in-app or chat message that names the observed pattern and the product that fits it, with a concrete comparison drawn from the customer's own account; email is the fallback for customers not in the app.
  4. 4
    Follow-through. the response is written back to memory and sets a cool-down; if the account behavior changes before the send (the balance moves, the transfers stop), the stale message is cancelled. Opening the product happens in the bank's own onboarding flow — the agent hands off, it never executes the signup itself.
Example
AgentYou've kept a sizable balance in your everyday account for a while now. Our flexible savings account would have been earning interest on it the whole time, with no lock-in. Want to see what that looks like on your numbers?
CustomerWhat's the catch?
AgentNo catch and no lock-in — you can move money back any time. Here's the comparison based on your recent balance. If it fits, I'll take you to the opening screen in your app and the bank's own flow handles the rest.

Configuration

How the agent is wired for this use case.

Triggera core-banking account signal fires — a balance crossing an idle threshold, recurring foreign-currency transfers, a salary-deposit change, or repeated card spend in a category another product serves better.
Tools & actions
  • Core banking event stream · receives the account signal that starts the flow
  • CRM · reads products held, eligibility and suitability flags, and the customer's offer/decline memory
  • In-app messaging surface · composes the recommendation naming the observed pattern with a comparison from the customer's own account
  • Email channel · sends the same recommendation as a fallback for customers not in the app
  • Onboarding flow (bank-owned) · receives the handoff to open the product; the agent never executes the signup
Autonomysignal scoring, message composition, and send run unattended, judge-gated, with a judge gating every unprompted message; suitability and eligibility rules are enforced before any offer is composed and a regulated-communications policy governs every send, while product opening always routes to the bank's own flow.
Channelsin-app · chat · email
Escalationopening the product hands off to the bank's own onboarding flow; if account behavior changes before the send, the stale message is cancelled and a decline sets a cool-down.

What you need

The inputs this use case runs on. Your channels stay yours; the agent supplies the judgment.

Signals

core-banking events (deposits, balance thresholds, recurring transfers, card-spend categories), product-usage events, life-event indicators the bank already models

Data

products held, eligibility and suitability flags, consent state, customer memory of past offers and declines

Guardrails

suitability and eligibility rules enforced before any offer is composed; judge gating on every unprompted message; frequency caps and decline cool-downs; regulated-communications policy; the bank keeps owning the send channels

Metrics it moves

  • attach-rateup, as products-per-customer grows when the offer lands at the signal moment instead of in a quarterly campaign
  • arpuup, because each attached product adds revenue from a customer the bank already acquired
  • ltvup, since multi-product customers are harder to displace and stay longer

See it on your own customer journey

Bring one drop-off, one churn cliff, or one silent segment. We will show you what a proactive agent with memory and judgment does with it.

Book a demo