CRM and loyalty: knowing each customer and writing to them at the right time, without handling files.
The CRM manager keeps the customer base and brings it to life: segmentation, e-mail and SMS campaigns, customer journeys, loyalty programme, consents. The data comes from the tills, the online store, the boutiques and customer service, and it rarely agrees. Automation reconciles it and triggers the messages; the team chooses whom to write to and what to say.
The role
Between marketing, the boutiques, e-commerce and customer service.
CRM turns customers’ purchases and interactions into usable knowledge: who buys what, how often, through which channel, with what consent. It designs the campaigns and journeys that sustain the relationship, runs the loyalty programme and measures what each action brings in. It also looks after the quality of the base and GDPR compliance, with the compliance officer.
Day to day
A typical day.
- MorningCheck the previous day’s sends: deliverability, opens, clicks, unsubscribes.
- MorningPrepare the targeting for the week’s campaigns from the segmentation.
- MiddayReview e-mail and SMS content with communication.
- AfternoonDevelop the customer journeys: welcome, post-purchase, birthday, reminder.
- AfternoonAnswer the boutiques and customer service on points, benefits and duplicate records.
- End of dayTrack the base’s indicators: active customers, acquisition, consents, customer value.
The objectives
What the role is measured on.
- Active customers
- Repeat purchase rate
- Customer value
- Database quality
- Reachable contacts with valid consent
- Campaign profitability
Indicators are named, never quantified: levels depend on each company, and the diagnostic measures them.
Day to day
The problems that come up most.
Duplicate customer records
The tills, the website and the CRM each create their own record; the same customer gets two messages, or none.
Segments rebuilt by hand
Each campaign starts again from an extract and a spreadsheet, with rules that change from one person to the next.
Poorly tracked consents
The opt-in collected in the boutique or on the website does not always follow the record through to the sending tool.
A loyalty programme that is hard to explain
Points, tiers and benefits are calculated in one tool and displayed in another; the boutiques cannot answer customers.
Uncontrolled marketing pressure
Several teams write to the same customer with no overall view, until they unsubscribe.
An effect that is hard to prove
Without a control group, there is no way of knowing whether the customer would have bought anyway.
By sector
One role, several realities.
- Fashion and luxury
Clienteling in boutiques, customers known personally by their sales associate, travellers buying in several countries, event invitations and highly personalised communication.
- Retail
Loyalty card at the till, a large file fed by stores and the website, campaigns around key dates and reactivation of occasional customers.
- Cosmetics and beauty
Repeat purchases of skincare products, advice and routines, points-based loyalty programmes, samples and skin diagnostics to link to the customer record.
- B2B distribution and wholesale
Store or trade cards, targeted coupons and offers funded with suppliers, independent or member stores sharing the same base.
What changes
Automate the task, keep the judgement.
What gets automated
- Reconciliation of customer records from the tills, the website and the boutiques, with duplicates flagged.
- Calculation of the segmentation (RFM, value, life cycle) at every sales update.
- Triggering of customer journeys according to the customer’s behaviour and consent.
- Calculation of loyalty programme points, tiers and benefits, visible at the till and on the website.
- Measurement of each campaign against a control group, in a shared dashboard.
What stays human
- Choosing targets, offers and the tone of messages.
- The loyalty programme rules and their cost to the company.
- The GDPR framework, set with the compliance officer.
Use cases
How to automate, case by case.
Each case starts from a question leadership teams ask. The starting point, what gets automated, the tools involved, and what stays in the teams’ hands.
How do you automate RFM customer segmentation?
- Today
- Segmentation is redone before every campaign: extracting sales, calculating recency, frequency and monetary value in a spreadsheet, then importing the lists into the sending tool. Rules vary from one person to the next and segments are out of date by the following day.
- What gets automated
- Segmentation rules are written once with marketing: recency, frequency and monetary thresholds, value segments, life-cycle stages. The calculation runs automatically at every sales update, and segments are synchronised with the campaign tool. Moves from one segment to another are tracked and can trigger a journey.
- What stays human
- CRM and marketing set the segmentation rules and revise them when the business changes.
- CRM
- Tills
- Campaign tool
- Reporting tool (BI)
How do you automate e-mail campaigns?
- Today
- Each campaign takes an extract, targeting, a template to fill in, rendering tests, then a send scheduled by hand. Exclusions, such as customers who have just bought the product or those who have withdrawn consent, are handled case by case.
- What gets automated
- Campaigns start from a calendar and synchronised segments: targeting, exclusions and consent are applied automatically at the time of sending. Templates are filled from the product repository and approved visuals, and an AI agent proposes variants of subject lines and copy. Results flow into the dashboard with those of the other channels.
- What stays human
- The team approves the content, the targeting and the final sign-off of each campaign.
- Campaign tool
- CRM
- Product repository (PIM)
- AI agent
How do you automate welcome, post-purchase and birthday customer journeys?
- Today
- Welcome or birthday messages go out when someone remembers to extract the list. A customer who buys in a boutique does not get the same follow-up as one who bought online.
- What gets automated
- Each journey is described once: trigger event, delay, content, channel, exit conditions. Events arrive automatically from the CRM, the tills and the website, whatever the purchase channel, and the journey runs in the campaign tool while respecting consent. A dashboard tracks the effect of each journey on repeat purchases.
- What stays human
- CRM designs the journeys, chooses their offers and approves each template message.
- CRM
- Campaign tool
- Tills
- Reporting tool (BI)
How do you automate management of the loyalty programme?
- Today
- Points are calculated in one tool, benefits are applied at the till under another rule, and the customer cannot find their balance in the boutique. Complaints go to customer service, which recalculates by hand.
- What gets automated
- The programme rules, points, tiers, benefits, expiry dates, are kept in one place. The calculation is done at every purchase, whatever the channel, and the balance is visible at the till, on the website and in messages. Tier changes and points about to expire automatically trigger a notification to the customer.
- What stays human
- Management sets the rules and the cost of the programme; customer service handles individual complaints.
- Loyalty software
- CRM
- Tills
- Campaign tool
How do you automate tracking of GDPR consent and opt-ins between the boutique, the website and the CRM?
- Today
- Consent is collected on paper or at the till in the boutique, by tick box on the website, sometimes by customer service. It does not always follow the record through to the sending tool, and an unsubscribe made in one tool is not applied in the others.
- What gets automated
- Each consent is recorded with its date, source, purpose and channel, then synchronised to every tool that sends messages. Objection and erasure requests are applied everywhere and logged. A regular check flags records contacted without valid consent.
- What stays human
- The compliance officer sets the rules; CRM handles special cases and customer requests.
- CRM
- Consent management platform
- Campaign tool
- Tills
How do you automate deduplication of customer records between the tills, the website and the CRM?
- Today
- A customer created in the boutique with their phone number, then on the website with their e-mail, exists twice. Their purchases are split, their value is underestimated, and they may receive the same message twice.
- What gets automated
- Matching rules compare e-mail, phone, name, address and card number, with a similarity score. Certain merges are done automatically, keeping the history and the most recent consent; doubtful cases are put to the team. A check when a record is created, at the till and on the website, prevents new duplicates.
- What stays human
- The CRM team decides on doubtful matches and approves the merge rules.
- CRM
- Tills
- Data quality tool
- AI agent
How do you automate lists of customers for sales associates to contact?
- Today
- In the boutique, each sales associate keeps their own customer book, on paper or on their phone. They contact the ones they remember, and a loyal customer may no longer be contacted when their associate leaves.
- What gets automated
- Each day, the tool offers each associate a short list of customers to contact, with the reason: arrival of an awaited product, birthday, alteration ready, long absence. An AI agent prepares a suggested message, which the associate personalises and sends from the boutique app. The contact is recorded on the customer record.
- What stays human
- The associate chooses whom to contact, how and with what message; the relationship remains theirs.
- CRM
- Clienteling app
- AI agent
- Tills
How do you automate detection of customers at risk of not coming back?
- Today
- Customers who drift away are spotted after the fact, when they have already been inactive for a long time. Reminders go out in bulk, without taking each customer’s usual buying rhythm into account.
- What gets automated
- The tool calculates an expected buying rhythm for each customer, from their history and that of comparable customers, and flags those who deviate from it. These customers enter a journey suited to their value: a message, attention from the associate or an offer. The effect is measured against a control group.
- What stays human
- CRM chooses the actions and offers made at each value level.
- CRM
- Reporting tool (BI)
- Campaign tool
- AI agent
How do you automate measuring the effect of CRM campaigns with a control group?
- Today
- Campaigns are judged on opens, clicks and sales from the targeted customers. Nobody knows what these customers would have bought without the campaign, and every action looks profitable.
- What gets automated
- For each campaign or journey, a share of the targeted customers is drawn at random and excluded from the send. The tool then compares purchases by the two groups, across all channels, and calculates the difference, its reliability and the associated cost. Results are kept by campaign type to guide the next ones.
- What stays human
- CRM and marketing decide which campaigns to continue or stop based on this measurement.
- CRM
- Campaign tool
- Tills
- Reporting tool (BI)
How do you automate management of marketing pressure per customer?
- Today
- CRM, e-commerce, the boutiques and sometimes the group’s other brands write to the same customers without coordinating. Some receive several messages in the same week and unsubscribe, others receive nothing.
- What gets automated
- Pressure rules are set by channel and by segment: maximum number of messages over a period, priorities between campaigns, rest periods after a purchase. Before each send, the tool checks these rules customer by customer and removes those who have reached their limit, in favour of priority messages. A dashboard shows actual pressure by segment and its link with unsubscribes.
- What stays human
- CRM sets the caps and decides priorities between teams.
- Campaign tool
- CRM
- Reporting tool (BI)
Questions from the teams
What teams look to improve.
- How do you automate a brand’s CRM?
- By first automating what feeds the base, then what comes out of it. Records arrive reconciled from the tills and the website with their consent, segments recalculate themselves, and campaigns and journeys go out from the sending tool. The team focuses on targeting and messages.
- What is the difference between a CRM and a campaign tool?
- The CRM holds knowledge of the customer: record, purchase history, preferences, consent. The campaign tool sends e-mails and SMS and measures responses. The two must exchange data continuously, or targeting runs on outdated data.
- Can AI personalise the messages sent to customers?
- It can propose product recommendations, adapt approved content to a segment and choose the sending time based on history. Template messages, offers and the limits of personalisation are set and approved by the team.
- Which indicators should be tracked to run a CRM?
- A few stable indicators are enough: active customers, acquisition, repeat purchase rate, customer value, share of reachable contacts with consent, measured effect of campaigns. They are calculated automatically from the CRM and the tills, with the same rule every period.
- How do you connect the tills to the CRM?
- Through an automatic flow that sends each receipt linked to an identified customer, with its products and its store. Quality depends on identification at the till: loyalty card, e-mail or phone number, entered consistently by sales staff.
- Do you need to change CRM to automate loyalty?
- Rarely. Automation plugs into the CRM and campaign tool in place and fills the gaps between them. The diagnostic shows whether a limitation of the tool justifies replacing it.
- How long should an inactive customer’s data be kept?
- The period is set with the compliance officer, in line with GDPR and the supervisory authority’s recommendations on prospecting. Once the rule is written, automation spots the records that reach it, sends a final request if one is planned, then anonymises or deletes.
- Is the data in the customer base protected?
- Access is limited by role, personal data is processed in line with GDPR, and hosting is chosen with the IT department. AI tools receive only the data needed for the task.
Related roles
02 — Getting started
Describe what’s holding your teams back.
A first conversation to qualify your need. Then a scoped diagnostic: the scope, the solution, the estimate.
Discuss your projectReply within 48 business hours.