CRM Strategy
Consent, Preferences and Eligibility: Who Can You Message?
Separate messaging permissions, customer preferences and offer eligibility, with a practical framework for regulated and cross-channel teams.
A practical way to organise messaging permissions, preferences and eligibility across email, SMS, push and WhatsApp.
TL;DR
- The Three-Layer Rule: For marketing, check channel permission and the lawful basis, customer preferences, and offer eligibility before sending. Mandatory service and regulatory notices need a separately reviewed process.
- Consent ≠ Preference: Having legal permission to message someone does not mean they want that message, on that channel, at that time. Conflating the two destroys trust and increases complaints.
- Eligibility is the hidden layer: Even with consent and preference satisfied, regulatory, risk, and operational constraints can make a message illegal, harmful, or irrelevant.
- Order matters: Document your Consent → Preference → Eligibility checks. The order is an operating choice, not a compliance guarantee.
- Implementation is data + rules + governance: You need a consent ledger, preference profiles, eligibility flags, and a decisioning layer that enforces the gating logic before any send.
- Marketing Ops owns orchestration; Compliance owns audit: Clear RACI prevents gaps—Marketing cannot self-certify eligibility; Compliance cannot slow-ship every campaign.
- Measure performance: Track complaints and engagement after implementing controls; outcomes are not guaranteed.
The Framework in 60 Seconds
The examples below are an operating framework, not legal advice. Rules depend on jurisdiction, channel, purpose and message content. A service notice may rely on a lawful basis other than consent; required notices must not be blocked by promotional caps. Ask legal and privacy specialists to approve the implementation, especially where sensitive customer data is involved. See the European Data Protection Board on lawful bases and the UK ICO direct marketing guidance.
- Define Consent: Capture and store legal permission for each channel + purpose (marketing vs. service), with timestamp, source, and proof.
- Capture Preference: Let customers declare topics, cadence, channels, and quiet hours—and honor these even when consent allows more.
- Maintain Eligibility Flags: Track product state, underwriting status, arrears, KYC, age, jurisdiction, vulnerability, and capacity constraints in real-time.
- Apply the Three-Layer Gate: Before any message, check Consent → Preference → Eligibility in order. Only messages passing all three gates enter the candidate set.
- Enforce Frequency Caps and Priority: Among candidates, apply caps and priority ordering (service > risk > lifecycle > promotions).
- Audit and Iterate: Log every decision for compliance, monitor suppression rates, and refine rules based on outcomes.
The Problem: Why Regulated Messaging Breaks
Messaging in regulated industries—finance, insurance, health, telecom—fails in two directions:
Risk Direction
Messages sent without proper consent, to ineligible customers, or during prohibited windows create regulatory exposure, customer complaints, and brand damage. A cross-sell offer to a customer with an open claim, a promotional SMS during a cooling-off period, or a product offer to someone who failed underwriting—these aren't just bad experiences; they can trigger regulatory action.
Performance Direction
Over-suppression or generic messaging destroys engagement. When marketing cannot distinguish between "cannot message" and "should not message this content," they either blast everyone (risk) or message no one (waste). Teams without clear rules oscillate between both failure modes.
The root cause is treating messaging as a single gate: "Do we have consent?" But regulated messaging requires three distinct layers, each with different owners, data sources, and failure modes.
Definitions: Consent vs. Preference vs. Eligibility
Consent
What it is: Where consent is required or chosen, record the person's specific choice for the channel and purpose, with its source, timing and evidence. Consent is not the only possible lawful basis for processing; applicable channel rules must be checked separately.
Components:
- Channel (email, SMS, push, WhatsApp, phone)
- Purpose (marketing, service, transactional, research)
- Timestamp (when granted)
- Source (web form, app, call center, paper)
- Proof (audit trail, recording reference)
What it is NOT:
- Consent is not permanent—it can be withdrawn
- Consent is not transferable—consent for email does not grant SMS consent
- Consent is not unlimited—marketing consent does not cover all content types
- Consent is not preference—having consent does not mean the customer wants the message
Preference
What it is: Customer-stated choices about how, when, and what they want to receive—even within the boundaries of consent. Preferences are typically stricter than consent.
Components:
- Topics/categories (product updates, offers, tips, research)
- Cadence (daily, weekly, monthly, only important)
- Channels (prefer email over SMS, no phone calls)
- Quiet hours (no messages before 9am or after 8pm)
- Language preference
What it is NOT:
- Preference is not consent—customers cannot preference their way into a channel they haven't consented to
- Preference is not targeting—your segmentation logic is not the same as customer-stated preference
- Preference does not override eligibility—preferring to receive offers doesn't make an ineligible offer legal
Eligibility
What it is: Business, regulatory, and risk constraints that determine whether a specific message, offer, or action is allowed for a specific customer at a specific moment.
Components:
- Product state (active policy, pending application, lapsed, in claims)
- Underwriting state (approved, pending, declined, referred)
- Financial state (current, arrears, collections, bankruptcy)
- KYC/AML status (verified, pending, flagged)
- Age/jurisdiction restrictions (product not available in region, age-restricted)
- Cooling-off periods (for example, a firm may choose to restrict cross-sell within X days; check the applicable rules)
- Vulnerability flags (financial difficulty, bereavement, health crisis)
- Capacity constraints (call center cannot handle inbound from this offer)
What it is NOT:
- Eligibility is not consent—a customer can be eligible for an offer but have no marketing consent
- Eligibility is not targeting—qualifying for a segment does not mean eligible to receive the offer
- Eligibility is not static—it changes with customer state, time, and context
The Three-Layer Rule: Decision Order
The Three-Layer Rule is simple: A promotional message should pass the applicable permission, preference and eligibility checks. Required service notices follow a separate approved path.
┌─────────────────────────────────────────────────────────────────────────┐ │ THREE-LAYER GATING FUNNEL │ ├─────────────────────────────────────────────────────────────────────────┤ │ │ │ [Target Audience] │ │ │ │ │ ▼ │ │ ┌──────────────┐ │ │ │ CONSENT GATE │ ─── Does legal permission exist for this │ │ │ │ channel + purpose? │ │ └──────┬───────┘ │ │ │ Pass │ │ ▼ │ │ ┌──────────────┐ │ │ │PREFERENCE │ ─── Does customer preference permit this │ │ │GATE │ topic + cadence + time window? │ │ └──────┬───────┘ │ │ │ Pass │ │ ▼ │ │ ┌──────────────┐ │ │ │ELIGIBILITY │ ─── Do policy, risk, and operational │ │ │GATE │ constraints permit this offer/action? │ │ └──────┬───────┘ │ │ │ Pass │ │ ▼ │ │ [Message Candidate Set] │ │ │ │ │ ▼ │ │ ┌──────────────┐ │ │ │ FREQUENCY & │ ─── Apply caps, priority ordering, │ │ │ PRIORITY │ conflict resolution │ │ └──────┬───────┘ │ │ │ │ │ ▼ │ │ [SEND / NO SEND] │ │ │ └─────────────────────────────────────────────────────────────────────────┘
Why Order Matters
- Efficiency: Checking permission first may reduce later lookups; measure the effect in your system.
- Compliance: Documenting checks supports review, but sequence alone does not demonstrate compliance.
- Clarity: Each layer has a different owner and data source. Mixing them creates accountability gaps.
- Relevance: Passing the checks can support a more relevant message, but does not prove legality or that the customer wants it.
Decisioning Logic: Suppression, Caps, and Priority
Once messages pass the three-layer gate, additional logic determines what actually gets sent.
Suppression Rules
Suppression rules remove messages from the candidate set based on context:
- Claims open: Suppress cross-sell and promotional messages
- Arrears/collections: Suppress all non-service messages
- Vulnerability flag: Suppress all sales messages; allow service only
- Underwriting pending: Suppress offers for the pending product
- Recent purchase: Suppress cross-sell for cooling-off period
- Complaint open: Suppress all marketing; elevate to resolution team
Frequency Caps
The following are illustrative promotional caps, not legal limits. Do not suppress required notices:
- Channel caps: Maximum 1 SMS per week, 3 emails per week
- Category caps: Maximum 1 promotional message per 14 days
- Global cap: Maximum 5 touches per customer per week across all channels
Priority Ordering
When multiple messages qualify for the same customer in the same window:
- Critical service (payment confirmation, security alert)
- Risk/compliance (regulatory notice, policy lapse warning)
- Service/transactional (renewal reminder, claim update)
- Lifecycle (onboarding, re-engagement)
- Promotional (cross-sell, upsell, offers)
Lower-priority messages defer to higher-priority messages within the same window.
Conflict Resolution
When two messages of equal priority compete:
- Recency rule: Favor the message triggered most recently
- Relevance score: Favor higher-predicted engagement
- Value rule: Favor higher expected LTV impact
Table 1: Three-Layer Examples Library
The table is illustrative, not a statement of universal legal requirements. Assess each notice, legal basis and channel separately; the terms ‘service consent’ and ‘mandatory’ below represent hypothetical internal policy, not general law.
| Scenario | Consent Req. | Preference Req. | Eligibility Req. | Notes |
|---|---|---|---|---|
| Service notification | Service consent for channel | N/A—service overrides | Active policyholder | No marketing consent needed |
| Quote follow-up | Marketing consent | Topic permits product updates | Quote not expired; no decline | May be transactional in some cases |
| Policy renewal | Service (factual) or marketing (upsell) | Cadence + quiet hours | Renewable; no lapse pending | Classify based on content |
| Missed payment / arrears | Service consent only | N/A | In arrears; not in collections | Promotional suppression mandatory |
| Underwriting pending | Service for updates | N/A for updates | Suppress same-product offers | May offer different products |
| Claims in progress | Service for claim updates | N/A for updates | Suppress all cross-sell | Promotional contact damages trust |
| Age-restricted product | Marketing consent | Topic permits offers | Age verified; jurisdiction permits | Cannot estimate age |
| Vulnerability flag | Service consent only | N/A | Suppress all sales | Human review before contact |
Implementation Blueprint
Data Requirements
Consent Ledger:
- Customer ID, Channel, Purpose, Status (granted/withdrawn)
- Timestamp (grant/withdraw date), Source, Proof reference
Preference Profile:
- Topics opted in/out, Cadence preference, Channel preference
- Quiet hours (start time, end time, timezone), Language preference
Eligibility Flags:
- Product states, Underwriting states, Financial state
- KYC/AML status, Age, Jurisdiction, Vulnerability flag
- Cooling-off windows, Capacity flags
System Touchpoints
- CRM/Lifecycle Tool: The execution layer—receives the final candidate set and sends messages.
- Consent Management / Preference Center: The capture layer—records and updates consent and preference.
- Data Warehouse / CDP: The consolidation layer—aggregates eligibility flags from source systems.
- Decisioning Layer / Rules Engine: The gating logic—applies the three-layer rule, suppression, caps, and priority.
Minimum Viable Implementation
For teams without heavy tooling:
- Rules table: A spreadsheet or simple database table defining suppression rules, caps, and priority
- Suppression list: Daily-updated list of customers who should not receive specific message types
- Preference center: Even a simple web form that updates a database table
- Manual audits: Weekly review of a sample of sends to verify rule compliance
This is not ideal, but it is better than no gating at all.
Table 2: Governance RACI for Regulated Messaging
| Responsibility | Marketing | Mktg Ops | Product | Data | Compliance | CX |
|---|---|---|---|---|---|---|
| Consent capture | C | A | R | C | R | C |
| Preference center | R | A | R | C | C | C |
| Eligibility rules | I | C | C | C | A | I |
| Suppression logic | C | R | C | A | C | I |
| Audit & incidents | I | C | I | C | A | C |
| Content approvals | R | C | I | I | A | I |
Legend: R = Responsible, A = Accountable, C = Consulted, I = Informed
Experimentation: Safe Testing in Regulated Contexts
Incrementality testing is valuable but must respect the three-layer rule.
Rules for Safe Testing
- Test only within eligible and consented audiences: Never include customers without consent or eligibility in test groups.
- Never hold out critical service or risk messages: Holdouts apply to marketing and lifecycle messages only.
- Maintain audit trails: Log which customers were in test vs. control, what message they would have received, and what they actually received.
- Use stratified randomization: Ensure test and control groups are balanced on eligibility factors.
- Analyze incremental outcomes: Measure incremental conversions, revenue, and complaints/regrets.
Why This Improves Personalization
Testing within the three-layer framework produces cleaner signals. You measure the effect of the message, not the effect of messaging ineligible customers. You build a library of what works within compliance—not what works if you ignore compliance.
Example 1: Insurance — Quote Follow-Up with WhatsApp
Scenario: A motor insurance prospect starts a quote, provides phone number with marketing consent for SMS and WhatsApp, but does not complete the purchase.
Three-Layer Check
✓ Consent Gate
Marketing consent for WhatsApp exists. Pass.
✓ Preference Gate
No opt-out of product reminders. Send time is outside quiet hours. Pass.
✓ Eligibility Gate
- Quote still valid? ✓ (7 more days)
- Already converted? ✓ No active policy
- Declined application? ✓ No
- Vulnerability flag? ✓ No
Pass.
Final Decision: Send WhatsApp follow-up.
What Changes If Underwriting Is Pending?
Eligibility Gate would fail for the same product. A different message ("We're reviewing your application") could pass as a service message under service consent.
Example 2: Banking — Payment Reminder vs. Cross-Sell Conflict
Scenario: A credit card customer has missed a payment (3 days overdue). The customer is also eligible for a personal loan cross-sell based on their profile.
Payment Reminder Check
Consent: Service consent for email ✓
Preference: N/A for service messages
Eligibility: Customer in arrears (not formal collections) ✓
Final: Send payment reminder.
Personal Loan Offer Check
Consent: Marketing consent for email ✓
Preference: No opt-out of product offers ✓
Eligibility: Customer in arrears → SUPPRESSION RULE TRIGGERED
Final: Suppress personal loan offer.
Why This Matters
Without the eligibility gate, the customer would receive both messages: a payment reminder AND a loan offer. This is legally risky (responsible lending rules), reputationally damaging, and operationally wasteful. The three-layer rule prevents this automatically.
Common Failure Modes and Prevention
| Failure Mode | Prevention |
|---|---|
| Assuming consent implies preference | Separate consent and preference data; honor preference even when consent allows more |
| Channel consent captured, purpose not | Capture purpose at consent point; default to most restrictive interpretation |
| Eligibility not checked in real-time | Real-time eligibility lookup for high-risk offers; suppress if uncertain |
| Preference center too coarse | Granular topic and cadence options; make it easy to prefer less without opting out |
| Vulnerability flags not operationalized | Explicit suppression rules for every flag type; test with synthetic data |
| No cooling-off enforcement | Cooling-off windows as explicit eligibility rules with countdown timers |
| Capacity constraints ignored | Capacity flags as eligibility layer; coordinate campaign timing with operations |
| Audit trail gaps | Log every gate decision with timestamp and data snapshot |
FAQ
1. What is the Three-Layer Rule for messaging?
The Three-Layer Rule states that any message must pass three gates in order: Consent (legal permission), Preference (customer choice), and Eligibility (business/regulatory constraints). Only messages passing all three gates should be sent. This can support a review of relevance and requirements, but it does not ensure compliance.
2. What is the difference between consent and preference?
Consent is legal permission—a binary, timestamped record that proves you can legally use a channel for a purpose. Preference is what the customer actually wants within that permission. A customer may consent to email marketing but prefer only monthly newsletters. Honoring preference builds trust and engagement.
3. What is eligibility in messaging?
Eligibility refers to business, regulatory, and risk constraints that determine whether a specific message or offer is appropriate for a specific customer at a specific moment. Examples include product suitability, underwriting state, financial state, vulnerability flags, and cooling-off periods.
4. Who owns consent, preference, and eligibility?
Consent capture is typically owned by Product with Compliance oversight. Preference center design is owned by Marketing Ops. Eligibility rules are owned by Compliance/Legal with input from Product, Risk, and Operations. Marketing Ops owns the orchestration that applies all three layers.
5. How do I handle service vs. marketing messages?
Assess the actual content and purpose, the applicable lawful basis and channel rules. A genuinely necessary service notice may use a basis other than consent; adding marketing content can change its classification. Respect marketing opt-outs and document uncertain cases for legal review.
6. What should I suppress during an open claim?
Suppress all cross-sell and promotional messages during an open claim. Allow service messages related to the claim. Customers in claims are sensitive—promotional contact damages trust at a critical moment.
7. How do I handle vulnerability flags?
Vulnerability flags should trigger suppression of all sales and promotional messages. Service messages should be reviewed for tone and necessity. Some firms require human approval before any outbound contact. Flags should have review dates—vulnerability is often temporary.
8. Can I do incrementality testing under the Three-Layer Rule?
Yes, but test and control groups must both be drawn from consented, preference-compliant, eligible audiences. Never hold out critical service messages. Maintain full audit trails. Testing within the framework produces cleaner signals.
9. How do I implement this without a sophisticated tech stack?
Start with a rules table, a daily-updated suppression list, a simple preference center (even a web form), and weekly manual audits. This is not ideal but is far better than no gating. The framework is the same; only the implementation speed changes.
10. How does this framework improve marketing performance?
A carefully implemented rule set can reduce unwanted or inappropriate sends. Measure complaints, engagement and deliverability to see what changes; the result is not guaranteed.
30/60/90-Day Rollout Plan
Days 1–30: Foundation
- Audit current consent capture points and data quality
- Document all existing suppression rules (formal and informal)
- Create draft eligibility flag inventory from source systems
- Design preference center requirements
- Define RACI for ongoing governance
Success: Clear inventory of gaps; stakeholder alignment
Days 31–60: Implementation
- Implement consent ledger with full audit trail
- Launch or upgrade preference center with granular options
- Build eligibility flag feed from core systems
- Implement suppression rules for top 5 high-risk scenarios
- Establish frequency caps and priority ordering
Success: Three-layer gating operational for at least one campaign type
Days 61–90: Governance and Optimization
- Launch weekly governance review meeting
- Audit sample of sends for rule compliance
- Measure suppression rate, complaint rate, engagement rate by layer
- Refine eligibility rules based on outcomes
- Document playbook for campaign launch checklist
Success: Governance rhythm established; measurable improvement in complaint rates
Closing
The Three-Layer Rule is not a compliance burden—it is a performance enabler. By ensuring every message is legally permitted, customer-wanted, and contextually appropriate, you create the foundation for personalization that actually works.
Agree the policy with the people responsible for privacy, compliance and customer operations, then test it with synthetic cases before sending.
Start with the audit. Build the gates. Run the governance. Measure the outcomes. The framework is simple. The discipline is hard. The results are worth it.