Customer loyalty: CRM, levers and measurement

What is customer loyalty?

Customer loyalty refers to all the actions that lead customers to buy again and to recommend. It aims for two effects: repeat purchases and preference. A loyal customer comes back, and chooses the brand even when an alternative exists.

Two dimensions are often distinguished. Behavioural loyalty shows in the data: purchase frequency, customer tenure, basket size. Attitudinal loyalty shows in opinions: satisfaction, attachment, intention to recommend.

The two do not always go together. A customer can buy again out of habit without being attached. They will then leave as soon as a competitor shows up.

What are the types of customer loyalty?

There is no official classification. A practical grid distinguishes four drivers, depending on what keeps the customer:

Driver	Example	Strength
Satisfaction	The product works, the service responds	Strong: it is the basis of everything else
Benefits	Points, discounts, status, priority access	Medium: a competitor can offer more
Relationship	Dedicated contact person, useful content, community	Strong if it brings real value
Constraint	Minimum term, switching cost	Weak: the customer leaves as soon as they can

Customer loyalty in B2B and B2C: the differences

In B2C, customer loyalty targets a large number of people. Purchases are frequent and small. It relies on segmentation, automation and rewards programmes. Consent governs every commercial message.

In B2B, customers are fewer and contracts longer. Several people share the decision. Loyalty therefore rests on the relationship with a contact person and the quality of the follow-up. It also means proving the value delivered on a regular basis: reviews, dashboards, review meetings.

A B2B customer is rarely lost over price. They are lost through silence, or through a poorly prepared change of contact person.

Why build loyalty? What the numbers really say

Two claims are repeated everywhere. The first says that acquiring a customer costs several times more than keeping one. The ratios quoted vary from one page to the next, and none is tied to an identifiable study. We do not repeat them.

The second is associated with Bain & Company. It states that a 5% increase in retention can raise profits “by up to 95%”. An article published by Bain on January 20, 2006 recalls it. This figure is dated. It describes a range observed in some industries, not a law that applies to your business.

Beyond the calculation, a loyal customer brings reviews, recommendations and product feedback. These effects also feed acquisition. Positive word of mouth lowers the cost of social ads or paid search campaigns.

CRM, the foundation of customer loyalty

CRM (Customer Relationship Management) refers to both an approach and a software tool. The approach organises the company around knowledge of the customer. The software centralises contacts, interactions, purchases and requests.

Marketing CRM refers to using this data to target and personalise actions: segments, email sequences, offers, follow-ups. Without reliable, shared data, each department speaks to the customer as if meeting them for the first time.

Data	Use in customer loyalty	Watch point
Identity and contact details	Contact, personalise	Updates, duplicates
Purchase history	Segment, recommend, follow up	Linking online and in-store purchases
Interactions and complaints	Detect dissatisfaction before the customer leaves	Systematic logging by customer service
Consents	Know who can receive what	Proof and date of each consent
Engagement	Spot customers who are drifting away	Reliability of open tracking

The GDPR governs this data. In France, the CNIL (the French data protection authority) publishes a reference framework on customer management. It provides for keeping a customer’s data for the duration of the relationship, then for three years after it ends. For a prospect, the period is three years after collection or the last contact (CNIL questions and answers, January 28, 2022). This framework is not binding, but any departure from it must be justified.

Segmentation, personalisation and lead scoring

Segmentation splits the database into groups that call for different actions. In retail, a simple and proven method is RFM segmentation: recency of the last purchase, frequency of purchases, monetary value spent. It distinguishes, for example, the best customers, customers to reactivate and one-time buyers.

Segment (example)	Signal	Action
Best customers	Recent, frequent purchases, high basket	Recognition, early access, review request
New customers	Recent first purchase	Onboarding, usage tips, second purchase
Customers drifting away	Former frequent customer, no purchase for a long time	Reactivation message, survey
One-time buyers	One old purchase	Targeted offer, then removal from the database if they stay inactive

Personalisation then adapts the message to the segment or to the person. It does not need to be spectacular. The right product, at the right time, with the first name spelled correctly, already makes a difference.

Lead scoring in B2B

Lead scoring assigns points to each prospect. It rates their profile: industry, company size, job role. It also rates their behaviour: pages viewed, content downloaded, demo request. Above a threshold, the prospect is handed over to sales.

The model is reviewed every quarter. Scores are then compared with the deals actually closed.

Legally, scoring and fine-grained segmentation count as profiling under the GDPR. Article 22 governs fully automated decisions that have significant effects on people. It grants a right to human intervention (CNIL fact sheet on profiling). A score that guides a salesperson is not an automated decision. A score that on its own excludes a customer from an offer can be one.

Which customer loyalty levers should you use?

Levers are chosen according to the loyalty driver you are targeting and the purchase cycle. A furniture maker does not have the same opportunities for contact as a bakery.

• Service quality: deadlines met, fast answers, complaints handled. It is the most profitable lever and the least spectacular.
• Onboarding new customers: the first weeks after the purchase often decide the second purchase.
• Useful communication: tips, new products, content. The main channel remains email marketing, with respect for consent.
• Loyalty programme: points, tiers, benefits. It rewards existing loyalty more than it creates it.
• Referral: a benefit for both the referrer and the referred customer turns satisfaction into acquisition.
• Listening: surveys, reviews, interviews. Then show what changed thanks to the feedback.

A loyalty programme collects a lot of data. The customer can retrieve part of it under the right to data portability. On October 14, 2025, the CNIL clarified that this includes the barcodes of the products purchased that are linked to the customer. The amount of discounts obtained is also part of it (CNIL, loyalty programmes).

Commercial follow-ups sent to customers follow the rules of direct marketing. In B2C, a commercial email requires consent. The exception covers products similar to those already purchased (CNIL, email marketing).

Which customer loyalty tools should you use?

Tools are best grouped by function rather than by brand. A small business can cover its basic needs with two or three of them.

Tool	Role in customer loyalty	Check before choosing
CRM	Centralise contacts, purchases, interactions and consents	Import of existing data, adoption by the teams
Email marketing and automation	Send sequences based on segment or behaviour	Connection to the CRM, unsubscribe management
Loyalty programme	Reward repeat purchases	Cost of the benefits, data portability
Surveys and NPS	Measure satisfaction and intention to recommend	Open question, follow-up on answers
Customer service (tickets, chat)	Handle requests and keep their history	Response time, link with the customer record
Customer reviews	Collect and publish feedback	Moderation and replies to negative reviews

A tool does not build loyalty on its own. Its value depends on the data that connects it to the others, and on how the teams use it.

How do you measure customer loyalty?

Four metrics cover the essentials. Choose a period consistent with your purchase cycle. A month suits a subscription, a year suits an equipment purchase.

Metric	Calculation	Question it answers
Retention rate	(customers at the end of the period minus new customers during the period) ÷ customers at the start of the period	How many customers did we keep?
Attrition rate (churn)	customers lost during the period ÷ customers at the start of the period	How many are we losing?
Customer lifetime value (LTV)	average margin per purchase × number of purchases over the length of the relationship	How much does a customer bring in over their lifetime?
Net Promoter Score	% of promoters minus % of detractors	Would our customers recommend us?

Take a fictional example. The average margin is €30 per order, with three orders a year. The relationship lasts four years on average. Customer lifetime value comes to €360 (30 × 3 × 4).

If acquiring a customer costs €120, each customer brings in three times their acquisition cost. If the relationship drops to one year, they bring in only €90. Acquisition then runs at a loss.

The Net Promoter Score rests on a single question, scored from 0 to 10: “How likely is it that you would recommend [company] to a friend or colleague?”

Scores of 9 and 10 are promoters, 7 and 8 passives, 0 to 6 detractors (Bain & Company method). The score alone says little. The open question that follows (“why this score?”) tells you what to fix.

To link these metrics to site and campaign data, see our web analytics guide. To improve the repeat purchase rate on the site itself, see our conversion rate optimisation guide.

Building a customer loyalty strategy in six steps

1. Measure the starting point: retention, churn, customer value and the share of revenue generated by existing customers.
2. Understand why customers leave: interview lost customers, read complaints, spot the moment they drift away.
3. Fix the pain points: a delay, an unreadable invoice, a service no one can reach. No programme makes up for a recurring flaw.
4. Segment the database: at least new customers, regular customers, customers drifting away.
5. Choose the levers: two or three per segment, with a quantified target and a review date.
6. Adjust: when possible, compare a group that receives the action with a control group that does not.

Customer loyalty is part of the overall digital strategy. It shares data, channels and messages with acquisition.