News publishers often tell me about their retention struggles: “We do so much, but things don’t add up.” Onboarding journeys exist. Engagement e-mails go out. Winback flows run. Yet churn remains stubborn.

Analysts from Mather, an U.S. subscription intelligence company, observed recently that “subscription performance doesn’t behave like a set of independent KPIs but as a single system driven by the tight linkage of acquisition, engagement, retention, paywalls, pricing, and onboarding.”

That framing explains why retention so often disappoints when it is treated as a standalone programme owned by one person or a team.

What distinguishes publishers that make progress is not more activity, but a shift in how they think about the lifecycle itself. 

Retention does not start at renewal. It is the accumulated outcome of decisions made much earlier, often before a reader ever becomes a subscriber.

Metaphors matter. This is where linear funnel thinking about consumer decision-making breaks down. 

Funnels encourage local optimisation

Marketing pushes discounted offers. Conversion teams tighten paywalls. The newsroom focuses on engagement. 

Each move can look successful in isolation. The cost appears later, when subscribers churn because expectations set early were never matched by habit or perceived value.

Many leading publishers such as The Washington Post are replacing funnels with loop-based models such as flywheels or cyclones. In practice, this means thinking in two reinforcing loops: 

• One loop focuses on acquisition and activation, moving readers from anonymous to registered to paid by reducing friction and increasing relevance. 

• The second loop focuses on retention, turning paid subscribers into habitual, loyal users and eventually advocates.

From funnel stages to systems

Once you adopt this system view, the priorities for your work on the retention flywheel become clear:

• The first step is to define the outcome before designing tactics. Some publishers explicitly prioritise reducing early churn. Some focus on maximising lifetime value. Others build more “forever subscribers.” Without choosing one primary outcome, based on your business maturity, lifecycle work turns into a long list of initiatives without leverage.

• Next, define lifecycle stages with explicit entry and exit rules. For example, activation might mean four or more active days in the first 30 days, downloading an app, signing up for a newsletter. Risk might be defined by a drop below a usage threshold or a failed payment attempt. Publishers using digital experience platforms codify these rules so the system reacts automatically rather than relying on manual campaigns.

• Then focus on behaviours that actually correlate with retention. Across INMA benchmarks and case studies, frequency of usage consistently outperforms raw intensity. Many publishers now design their lifecycle around getting subscribers to four or more active days per month as early as possible.

• Before adding further sophistication, build a minimum viable lifecycle. Structured onboarding, habit nudges, payment defence, cancellation save, and winback still do most of the work. For example, Newsday reduced early churn by front-loading engagement while subscribers were still in “transaction mode.” Público improved retention simply by strengthening payment recovery and extending grace periods, a reminder that not all churn is about content.

• Concentrate effort on the first 100 days. Benchmarks clearly show most new subscribers churn within the first three months. Leading publishers deliberately structure this period, from reassurance in the first week, through habit formation in the first month, to reinforcement of routines before day 100. Mediahuis found once habit is established, renewal rates improve regardless of content mix.

• Only after this foundation is stable does it make sense to add AI and automation. Predictive churn models built in-house or rented from vendors help prioritise interventions. Most sophisticated decisioning tools align paywalls, engagement, and retention rather than treating them separately (see a case study of Hearst). What AI will not fix is unclear strategy or fragmented ownership between editorial, product, and marketing.

• Operationalise and review quarterly. Lifecycle work fails without ownership, operational level agreements between teams involved in subscriptions, and a learning cadence.   

Mather’s analysis reinforces the same lesson from a different angle. Fixing one KPI at a time produces local wins and delayed trade-offs. Retention reflects how well paywalls, pricing, onboarding, engagement, and renewal decisions are aligned across the system.

For retention leaders, the takeaway is straightforward. If retention feels harder than it should, stop asking which campaign to add next. Ask where the lifecycle system is misaligned.

Want to go deeper? On February 12, I will join Mather’s Peter Doucette for a candid discussion about the tradeoffs in managing subscription lifecycles. Register for a free live-stream and a playbook. 

Greg’s Readers First newsletter is a public face of a revenue and media subscriptions initiative by INMA, outlined here. INMA members can subscribe here.

Header art: Adobe Stock TarikVision.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .