Full article
Created by Brenden O'Sullivan · Edited by Marc Woodhead · Reviewed by Marc Woodhead · Published 18 March 2026
Subscription businesses often face the dilemma that the most interesting signals, like wearable events or usage spikes, should not be activated in CRM due to sensitivity and expectation mismatches. This is crucial in 2026 as pressure to act grows. My view: if a segment cannot survive scrutiny from operations, legal, product and channel owners, it should stay out of campaign logic. Audience activation governance stops growth work becoming rework.
Signal baseline
Not every product-collected signal is fit for marketing. Separate signals at source: operational signals for service delivery, product usage signals for feature adoption, and sensitive health or inferred wellbeing signals. The third class requires stricter governance.
A recent strategy call illustrated this. We tested using richer wearable-derived signals for CRM retention versus segmenting on clear subscription events. The evidence favoured the latter due to unstable definitions and permissions. A proper customer data operating model distinguishes what can be seen from what should be activated, with ownership attached.
What is shifting
The boundary between product and marketing data is becoming commercially tempting and operationally porous. Subscription businesses have more event streams and faster routes into CRM and ad platforms, but labels like 'engaged' or 'at risk' can lose context in handover. Activation lineage, tracking a field's origin, transformation, permissions, and destinations, is now a practical necessity to prevent misuse.
A blunt ban on sensitive data is counterproductive. Instead, define options precisely: allow service communications but restrict promotional segmentation where sensitivity or consent is weak.
Who is affected
Stakeholders include CRM managers, data leads, platform teams, and commercial leadership. Costs appear as slower campaigns, uncertain approvals, and duplicated audiences.
For a subscription health app with wearable integrations, product might use lapsing behaviour for coaching, retention for win-back sequences, and paid media for lookalikes, each with different sensitivity tolerances. Implementation teams face consent or semantic drift when signals move across fragmented tooling like Braze, Salesforce Marketing Cloud or Meta. Consent-aware segmentation should be a release discipline, not a policy statement. DNA helps tie segment definitions to provenance, permissions and destination logic.
Signals implying health state, vulnerability, or diagnosis should generally stay out of promotional CRM segmentation unless clear expectation, permissioning, channel suitability and documented lineage are evidenced. A strategy that cannot survive contact with operations is branding copy.
Where signals should stay out
Keep health or usage signals out of CRM segmentation in these situations: when the signal is too sensitive for the message, such as using sleep quality for a discount; when the signal is inferential, like 'stress risk' modelled approximations; when permission separation is weak, with broad service acceptance but unclear marketing consent; and when lineage is poor, lacking reconstruction of field creation and approval.
An unresolved tension is that the same signal may be appropriate in-product but not in CRM, requiring channel-specific rules.
Actions and watchpoints
Build a decision model before launch. First, classify signals by sensitivity and intended use. Secondly, create channel-specific rules for email, push, paid audiences and in-app messaging. Thirdly, require a proof pack before go-live: source field, transformation logic, consent state, owner, destination mapping and segment purpose. Fourthly, design fall-back segments based on low-risk operational data.
DNA helps teams move from argument to execution with governed audience builds, visible provenance and traceable decisions. This trade-off sacrifices speculative uplift for faster approvals and cleaner releases.
Start with one live segment and map its signals. If defending its origin, permission and purpose is hesitant, tighten the rule set. For a practical review, contact DNA to map options and next moves.
If this is on your roadmap, DNA can help you run a controlled pilot, measure the outcome, and scale only when the evidence is clear.
