AGENT · LOYALTY BRIEF NOTTURNO

The loyalty programme brief reaches the retail team in the morning.

Loyalty Brief Notturno produces every morning the brief on the retailer's loyalty programme: active customer segments, redemption anomalies, personalisation opportunities. The retail marketing team and store managers have the full picture ready before the store opens.

02 · AGENT IN ACTION

Loyalty Brief Notturno at work.

Context

Why it exists.

Loyalty programmes at mid-to-large retailers manage millions of customers with usage patterns that vary by segment, geography, and seasonality. Redemption dynamics, clusters of customers who suddenly stop redeeming points, and activation opportunities in premium segments are hard to monitor manually on a daily basis.

What it does

How it works each night.

Loyalty Brief Notturno activates every night. It reads the customer's loyalty programme management system, identifies the patterns from the previous day, and compares against the weekly and monthly baseline. It produces the brief in three blocks: active segments (who is using loyalty, and how), redemption anomalies, and personalisation opportunities.

Supervision

The decision stays with marketing.

The retail marketing team and store managers decide the day's actions. The agent does not launch campaigns autonomously, does not modify the loyalty programme without approval.

03 WHO IT SERVES

The teams that move from reactive monitoring to proactive loyalty management.

Retail marketing team

Has the loyalty programme picture updated every morning, without manually accessing the management system. Campaign or segment activation decisions are based on fresh data from the previous day, not on weekly reports.

Brief 07:00 12 active segments
Gold segment +8% engagement
Silver segment redemption -3%
Anomaly · campaign 14 discarded
Monday action map

Store manager

The store manager of a physical outlet sees the active segments in their zone and any local anomalies. The brief can be configured by level (company-wide, by geographic area, by store) to match the retail network's needs.

Cross-channel · 7d 1,420 mentions
Site chat NPS 7.2
Reviews top theme: shipping
Social · IG sentiment +0.4
Brief to the brand owner

CRM manager

Sees the correlation between loyalty programme usage, repurchase, and at-risk churn segments. Reactivation campaigns trigger before customers leave the programme, not after.

Switch window · 7d 240 customers
High-value cluster A callback prio 1
Dedicated offer €8/month · 12 months
Engagement < threshold escalation to operator
Expected win-back rate: 22%
04 EXAMPLE OF A PROCESS

A redemption anomaly detected and handled in the morning.

The morning brief

At half past six, fresh overnight data.

For a fashion retailer with three hundred stores and four million customers in the loyalty programme, the brief arrives every morning at half past six in the marketing channel. Loyalty Brief Notturno worked overnight: it read the previous day's redemption data, segmented by area, by customer segment (Gold, Silver, Bronze), by product type purchased with points. It compared against the twelve-week baseline.

The three situations

Anomalous cluster, opportunity, positive signal.

The brief flags three situations. A cluster of Gold customers in a specific zone has not made any redemption for several weeks — an anomaly relative to the segment baseline. One thousand two hundred and thirty Gold customers have points expiring in the next seven days — an activation campaign opportunity. The Silver segment in two regions shows nineteen percent growth in loyalty usage versus the monthly average — a positive signal to build on.

The decisions

Marketing decides. The agent records.

The marketing team decides: technical check on the anomalous cluster's cards, launch of a mini email campaign for Gold customers with expiring points, driver analysis for the Silver growth. Decisions stay in the runtime registry for effectiveness analysis.

05 CONFIGURATION

Declarative rules from the customer's retail marketing team.

The rules of Loyalty Brief Notturno are declarative. The customer's retail marketing team defines in a readable format the loyalty programme segments, the redemption anomaly thresholds (e.g. absence of redemption for N weeks on the Gold segment), the baseline comparison metrics, and the brief format per distribution level (company-wide, by area, by store). Loyalty programme management system integration is built during delivery via a dedicated adapter.

SPEC SHEET
Language
TypeScript (Node.js)
LLM model
customer's choice: Anthropic, OpenAI, Mistral, open source models hosted internally, AWS Bedrock for a private model
Built-in controls used
pii-detector, tool-rate-limit
Native delivery channels
Slack, Telegram, WhatsApp, OpenAI-compatible HTTP
Scheduling
configurable per instance (typical overnight, brief ready at 06:00-07:00)
Loyalty programme management system integration
dedicated adapter built during delivery by the Exelab team
Retail CRM integration
dedicated adapter built during delivery by the Exelab team
Memory
persistent per instance, pgvector + PostgreSQL FTS
Registry
immutable, queryable with a standard SQL client
06 FREQUENTLY ASKED QUESTIONS

How Loyalty Brief Notturno works in detail.

Yes. The brief can be configured by distribution level: a consolidated company-wide version for the central marketing team, an area version for regional managers, a store-level version for store managers. Granularity is defined during configuration. The customer's loyalty programme management system must track data at store level for the local brief to be available.

Anomaly rules include seasonality as a comparison dimension: the baseline is calculated against the same period in previous years (or comparable weeks). A redemption dip typical of the summer period does not generate an alert if it falls within the configured seasonal variation. Cases that exceed the expected variation are flagged. Seasonal thresholds are defined during configuration with the marketing team.

Yes, as long as both programmes are tracked in the customer's loyalty management system. Each programme can have segments, anomaly thresholds, and comparison metrics configured separately. Coverage of multiple programmes is defined during discovery.

The typical pattern is 10-14 weeks. Discovery and loyalty programme mapping two weeks, segment and anomaly rule configuration three weeks, loyalty management system and CRM integration three weeks, testing with real data and hand-off to the marketing team two weeks. Duration depends on the complexity of the loyalty management system and the variety of segments.

From a 30-minute conversation to the squad in production.

A 30-45 minute conversation to understand how Loyalty Brief Notturno would configure to the customer's case. Loyalty programme management system, customer segments, granularity level of the brief.