How it works
How a transaction becomes revenue.
Six steps, one signal, and a feedback loop that makes each cycle sharper than the last. All of it within a second of the purchase, not in an overnight batch.
The transaction journey
- 01
The cardholder pays
A card is presented at a merchant point of sale, online or in person. Nothing about the payment experience changes — for the cardholder, the merchant or the acquirer.
- 02
Transaction data reaches ActtraQ
A copy of the authorization message reaches the platform through the issuer processor or network connector, within a second of the transaction clearing. ActtraQ sits alongside the payment flow, never inside it.
- Visa VOP
- e-global
- NeoNet
- Issuer processors
- 03
The model scores the moment
Historical spending patterns, merchant category, the bank's own engagement rules and known cardholder preferences are evaluated together to decide whether this transaction is worth acting on, and how.
- Spending history
- Merchant category
- Bank engagement rules
- Cardholder preferences
- 04
The platform triggers an action
A personalised offer, a cashback calculation, a points redemption option, an advertising placement — or several at once, because they all derive from the same signal.
- Personalised offer
- Cashback
- Points redemption
- Ad placement
- 05
The cardholder is reached
Delivery over the channel the cardholder actually uses — SMS, email, push or in-app. The message confirms the transaction, delivers the benefit, and cross-promotes what comes next.
- SMS
- Push
- 06
The loop closes
Engagement is captured, the model is reinforced by what actually happened, and ROI is attributed back to every party — the bank, the merchant and the advertiser.
Cross-issuer intelligence
Every bank benefits. No bank is exposed.
A single issuer sees only its own cardholders, which caps how good its models can ever get. ActtraQ learns from aggregated, de-identified transaction signals across many issuers — so each participating bank benefits from purchase and engagement patterns none of them could surface alone.
- No issuer's data is ever exposed
- Learning happens on aggregated, de-identified signals — not on raw cardholder records
- Model quality compounds with every institution that joins
- Each bank's own outputs and reporting stay entirely its own
Single-bank platforms see one portfolio. We see the shape of all of them — without any of them seeing each other.
Walk through it against your own transaction flow.
We will map the six steps onto your processor topology and show where the integration points sit.