Leaders Insights
Leaders Insights

Rester au meilleur niveau, un peu chaque jour.

DomainesMarketingDataFinanceIA
RessourcesApprendreTestOutilsBlogGlossaire
© 2026 Leaders Insights — Tous droits réservés.
Formations/Data in fashion/Governance, privacy and checks/Governing customer and clienteling data through the consent lifecycle
3/4+150 XP

Governance, privacy and checks

10Mapping fashion's privacy obligations from loyalty to fitting-room tech+15011Building a data governance operating model for a fashion house+15012Governing customer and clienteling data through the consent lifecycle+15013Running a privacy and governance audit before peak season+150

Governing customer and clienteling data through the consent lifecycle

# Governing customer and clienteling data through the consent lifecycle

A VIP shopper walks into a Burberry store in Milan. The associate opens a clienteling app and sees her name, her past purchases, her size, and a note: "prefers oversized trench, allergic to wool." That single screen pulls from three systems. And if this shopper unchecked "marketing emails" online last month, that note may now be a compliance breach, not a service feature.

This lesson shows you how consent, preferences, and purchase history should flow, expire, and get deleted across your systems without turning off personalization overnight.

The three systems you are governing

Before consent, get the plumbing straight. Fashion retailers typically run customer data across three layers:

  • CDP (Customer Data Platform): a system that unifies customer data from web, app, email, and stores into one profile. Think Segment, Salesforce Data Cloud, or Bloomreach.
  • CRM (Customer Relationship Management): the system of record for identity, contact details, and order history. Salesforce, Microsoft Dynamics.
  • Clienteling / store associate apps: mobile tools associates use on the shop floor. Salesforce Retail, Tulip, or custom apps.

Data flows outward from CRM and into the associate's hand. That last hop is where governance most often fails, because a human is reading it live.

CRM
Customer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.
Voir la définition complète →
CDPCDPA Customer Data Platform unifies customer data from all sources into persistent, actionable profiles that other systems can use.Voir la définition complète →

What "consent" actually covers

Not all customer data needs consent, and treating it as one bucket is a classic mistake.

Under the GDPR (General Data Protection Regulation), the EU law governing personal data, and the UK GDPR, you need a lawful basis to process data. The two you will meet most:

  • Consent: freely given, specific, opt-in. Required for marketing emails, SMS, and profiling for advertising.
  • Legitimate interest: processing you can justify without asking, if it does not override the person's rights. Order fulfillment and fraud checks often sit here.

Purchase history to complete an order sits on contract as a basis. Using that same history to predict what she will buy next and email her about it is profiling for marketing, which usually needs consent.

In the US there is no single federal law. The strongest state rule is the CCPA/CPRA (California Consumer Privacy Act, amended by the California Privacy Rights Act), which gives Californians the right to opt out of the sale or sharing of their data and to request deletion. As of 2026, roughly twenty US states have comprehensive privacy laws (Virginia, Colorado, Connecticut, Texas, and more). Treat this as an estimate and check current status, because new states pass laws each session.

The practical takeaway: a single "yes" checkbox is not enough. MapMapUsing software to automate repetitive marketing tasks and campaigns, enabling personalisation at scale across channels like email, web, and social.Voir la définition complète → each data use to its own consent or basis.

The consent lifecycle: capture, store, propagate, expire, purge

1. Capture

Consent must be granular and logged. When a shopper signs up in a Paris boutique, the associate's app should capture separate opt-ins:

  • Email marketing
  • SMS marketing
  • Style profiling (using purchase and browsing data to personalize)
  • Sharing with brand partners

Log each with a timestamp, the exact wording shown, the channel, and the version of the consent notice. If a regulator asks "prove she agreed," you need that record.

2. Store

Consent state should live in one authoritative place, usually a consent management platform (CMP) or a dedicated table in the CDPCDPA Customer Data Platform unifies customer data from all sources into persistent, actionable profiles that other systems can use.Voir la définition complète →. Never let each system keep its own copy of "yes/no" with no sync. That is how a customer who opted out on the website still gets a text from the store.

3. Propagate

This is the hard part. When consent changes, every downstream system must reflect it fast. A realistic target is near real time for marketing suppression and within 24 hours for analytical systems.

Here is a simplified consent record and a suppression check:

python
consent = {
    "customer_id": "C-88421",
    "email_marketing": False,   # opted out 2026-03-01
    "sms_marketing": True,
    "style_profiling": True,
    "updated_at": "2026-03-01T09:14:00Z"
}

def can_send(channel, consent):
    return consent.get(f"{channel}_marketing", False)

can_send("email", consent)  # False -> suppress
can_send("sms", consent)    # True  -> allowed

The clienteling app should call the same source. If the associate sees a "send lookbook by email" button, it must be greyed out for this customer.

4. Expire

Consent is not forever. Some regulators and many internal policies treat consent as stale after a period of inactivity (commonly cited as 12 to 36 months, a policy choice, not a fixed legal number in most cases). If a customer has not engaged in, say, 24 months, you re-permission (ask again) rather than keep emailing.

Preferences also decay. A "prefers slim fit" note from 2021 may be wrong. Tag preferences with a source and date so associates know how fresh a note is.

5. Purge

When someone exercises their right to erasure (the GDPR "right to be forgotten") or a CCPA deletion request, you must delete their personal data across all systems, including backups within a reasonable window, and including that clienteling note about the wool allergy.

Exceptions exist: you can retain what you need for legal or tax reasons (order records for accounting). So "delete" often means "delete marketing and profiling data, retain minimal transaction records under a documented retention rule."

For the official text on erasure obligations, see the EU's own guidance on the right to erasure.

Keeping personalization alive while honoring consent

The fear is that consent kills personalization. It does not, if you separate uses.

A customer can opt out of email marketing but still allow style profiling for in store service. In that case:

  • The associate still sees size and fit preferences.
  • No automated email goes out.
  • The CDPCDPA Customer Data Platform unifies customer data from all sources into persistent, actionable profiles that other systems can use.Voir la définition complète → can still segment her for store clienteling but excludes her from email campaigns.

This is why granular consent matters commercially. Bundle everything into one switch and a single opt-out wipes out your entire relationship. Split it, and you keep the service layer alive.

The checks and audits to run

Governance is only real if you test it. Run these regularly.

Suppression reconciliation

Pick a sample of customers who opted out. Confirm they received zero marketing across email, SMS, and push in the following 30 days. A simple metric:

Suppression accuracy = (opted-out customers with zero sends) / (total opted-out customers)

Worked example: you sampled 500 opted-out customers. 497 received nothing; 3 got an email. Accuracy = 497 / 500 = 99.4%. Anything below 100% is a breach candidate; investigate the 3.

Consent-to-basis mapping audit

List every data use (recommendation engine, lookbook emails, associate notes, partner sharing) and confirm each maps to a documented lawful basis. Any use with no mapping stops until fixed.

Deletion completeness test

Submit a test erasure request and trace it. Did the CDPCDPA Customer Data Platform unifies customer data from all sources into persistent, actionable profiles that other systems can use.Voir la définition complète →, CRMCRMCustomer Relationship Management: software and strategy to manage and analyse customer interactions throughout their lifecycle.Voir la définition complète →, and clienteling app all purge? Did the data reappear from a nightly batch feed the next morning? Reappearance is the most common failure.

Freshness and drift check

Flag preferences and consent records older than your policy threshold. Report the percentage of active profiles with stale consent.

Vérification des acquis

1. A shopper unchecks 'marketing emails' online, but a store associate's clienteling app still displays a note used to tailor promotional outreach to her. Why does this represent a compliance risk rather than good service?

2. Why is the 'last hop' from CRM/CDP into the associate's clienteling app described as where governance most often fails?

3. A retailer processes purchase history to fulfill an order and run fraud checks without asking the customer to opt in. Which lawful basis best justifies this, and why?

CHOIX MULTIPLES

4. Select ALL correct answers about the distinction between consent and legitimate interest as lawful bases under GDPR.

Sélectionnez toutes les réponses correctes.

CHOIX MULTIPLES

5. Select ALL correct answers about the three data layers a fashion retailer governs.

Sélectionnez toutes les réponses correctes.

Roles and accountability

Governance needs named owners, not committees.

  • Data Protection Officer (DPO): required under GDPR for many retailers processing data at scale. Owns compliance and is the regulator's contact.
  • Data owner (marketing or CRM lead): accountable for a data domain and its correct use.
  • Store operations: accountable for how associates handle what appears on the app.

The regulators enforcing this include national Data Protection Authorities in the EU (France's CNIL, Ireland's DPC, which oversees many large firms headquartered there), the ICO (Information Commissioner's Office) in the UK, and state Attorneys General plus the California Privacy Protection Agency in the US. GDPR fines can reachreachThe number of unique people exposed to your message in a given period. Unlike impressions, reach counts each person once, no matter how often they see it.Voir la définition complète → up to 4% of global annual revenue, a genuine board-level risk for a large fashion house.

Key takeaways

  • Split consent by use. Separate email, SMS, profiling, and partner sharing so one opt-out does not destroy the whole relationship. This protects both compliance and personalization.
  • One source of truth for consent, propagated everywhere. The clienteling app on the shop floor must read the same live consent state as the email engine. Copies that drift cause breaches.
  • Consent and preferences expire. Re-permission stale consent (a 24 month inactivity window is a common policy choice) and date-stamp every preference note so associates trust what they see.
  • Deletion must be complete and repeatable. Test erasure requests end to end, and confirm data does not reappear from batch feeds. Retain only what a documented legal basis requires.
  • Audit continuously. Track suppression accuracy, consent-to-basis mapping, deletion completeness, and freshness. Governance you do not measure is governance you do not have.

Précédent

Building a data governance operating model for a fashion house

Suivant

Running a privacy and governance audit before peak season