Loyalty

(LOY)

Points, tiers, rewards and gift cards that work the same in store and online. A customer earns at the register and redeems on the web, because both channels read the same balance.

Points Tiers Rewards Gift Cards Store Credit Referrals

Key Features

Points, tiers, rewards and gift cards that work the same in store and online.

Points and Earning

Earning rules by spend, product, category or campaign. Points post at the moment of sale, in any channel.

Tiers

Tier thresholds with their own benefits, pricing and earn multipliers. Promotion and demotion run on a schedule you set.

Rewards and Redemption

Redeem points for discounts, products or shipping. Redemption is validated against a single balance, so it cannot be double-spent.

Gift Cards

Issue, load, check and redeem gift cards. The same card works at the counter and at online checkout.

Store Credit

Credit from returns and goodwill, tracked per customer and usable anywhere they buy.

Referrals

Referral codes with attribution and reward rules for both sides, tracked back to the originating customer.

Kubrik OS — agent layer

Loyalty, assisted and operated by AI

Loyalty is where the AI can act on one customer at a time and see the result in their next transaction.

AI-assisted

What the AI advises on

Available on every plan, Starter included.

Identifies customers whose frequency is dropping while their balance sits unused
Recommends the reward that actually moves the next purchase, per segment rather than per campaign
Resolves points, tier and store credit across store and web into one balance the customer recognises

AI-operated

What the AI does under governance

Professional and Enterprise. Starter runs the AI in advisory mode only.

Adjusts tiers when the qualifying criteria are met, up and down
Issues points, rewards and store credit inside the configured limits
Launches and closes targeted campaigns against a defined segment

Governed tools an agent can call in this module

adjustTier() issueReward() grantStoreCredit() launchCampaign()

Autonomy is a setting, not a promise — choose the rung per role and per module

Suggest

The agent proposes. Nothing changes until a person acts on it.

Prepare

The agent builds the transaction and leaves it waiting for approval.

Execute below threshold

Acts on its own under a value or risk limit you set.

Execute by policy

Acts inside a written policy, every action fully audited.

Escalate

Anything outside the policy stops and goes to a person.

Same rules, either operator. An agent obeys the same permissions, validation and business rules a person does, and every action lands in the audit trail: who or what initiated it, the inputs, the rationale, the approvals, the resulting transaction and how to reverse it.