Loyalty built into operations,not bolted on.

It is a native module inside marql, not a third-party integration. Loyalty keeps its own database for isolation, and it is linked to your decisions and the Impact Ledger inside the platform — so there is nothing to sync, no webhook mapping, and no separate bill: the price is per location and the module is not metered on top of it.

By visits and recency, on your own transaction data. Champions have eight or more visits and came back within the last 30 days; Loyal have four to seven; Regular is the rest of the active base; New joined inside the last 30 days; At Risk have not visited for 30 days or more. Each segment also carries its share of revenue, which is what tells you whether the small groups are the valuable ones. Separately, a nightly job scores every member one to five on recency, frequency and money, and the assistant can answer against those scores.

Points, which members earn per euro spent, so spend drives the tier indirectly. The tier follows the lifetime points total — Silver at 500, Gold at 2,000, Platinum at 5,000 by default — and you change both the thresholds and each tier's earn multiplier. The four tier names are fixed. Points expiry and the reward catalogue are configured once for the programme rather than per tier.

Discount percentage, bonus points multiplier, and free item. You set the target segment, the offer, a cap on how many redemptions the campaign may hand out, and the dates. Redemptions and the discount they cost are counted as they happen. Birthday, new-member and win-back campaigns can also run on their own.

Part of the target segment — 15% by default — is held back and gets nothing, and the two groups are frozen at the first send so a re-send cannot reshuffle them. When the measurement window closes, 14 days by default, the arms are compared with a two-sample test at α = 0.10 and the campaign lands in the Impact Ledger as helped, no effect, worse or inconclusive. The drafting is a conversation: ask the assistant to win back a segment and it proposes the segment, the offer and the duration, with nothing written until you approve it.

Yes — repeat visits by acquisition week, W0 to W4, for the four most recent weekly cohorts. You can see where repeat-visit drop-off happens and cross-reference it with the segments to understand which cohorts are ageing into At Risk.