Netflix's 18-Hour Benchmark: How Proxy Metrics Predict SaaS Churn

Netflix expects roughly 18 to 19 hours of consumed content per user, per month. Below that line, churn risk climbs; above it, the subscriber almost certainly stays. That single number, laid out in ProductLed's guide to customer retention metrics, is what the site calls a proxy retention metric — an action-based measure that predicts future retention before churn data confirms it.

What are retention metrics, and which ones matter?

Retention measures how long customers stick around and actively engage. ProductLed splits it into two lenses. Core retention counts the percentage of users still active after a fixed period — of the users you acquired three months ago, how many remain in Month Three. Proxy retention tracks the core actions that support your model: daily usage of an outbound email platform, or hours of content watched. Because proxy metrics capture behavior today, they let teams predict behavior tomorrow, next week, and three months out. Your proxy metric will usually resemble your North Star metric, the single goal-based measure such as monthly active subscribers.

Which KPIs belong on the dashboard?

ProductLed names seven:

• Customer Retention Rate (CRR) — the percentage of customers continuing to use the product over a period.
• Customer Churn Rate — the inverse, the percentage who stop within a time frame.
• Lifetime Value (LTV) — total revenue a customer generates over the relationship.
• Net Promoter Score (NPS) — likelihood to recommend, rated 0 to 10.
• Customer Satisfaction Score (CSAT) — a satisfaction rating scale.
• Customer Effort Score (CES) — how much work customers must do to reach their outcome.
• Time to Value (TTV) — how quickly users realize the product's benefit; faster TTV raises engagement and retention.

ProductLed also recommends tracking a Product Engagement Score (PES), which gauges engagement strength and serves as a leading indicator of retention.

How do cohorts and retention curves work?

A cohort is a group of users who experience product value at roughly the same time. If 100 people adopt in Month Zero, that group becomes your unit of analysis; you track core and proxy metrics per cohort rather than blending all users together. A retention curve plots the average share of a cohort still active over time. One worked example in the guide shows a January signup cohort decaying almost to zero within six to seven months — a curve that never flattens. ProductLed calls this a signature of bad product-market fit: the company has solved acquisition but not the mechanics of keeping users. When the curve starts flattening by around month three, users are finding repeat value, and that is the point, per the guide, where you can hit the gas on growth spending.

Building the measurement stack requires cohort-level signup tracking. ProductLed names Google Analytics, Google Analytics 360, Segment, Mixpanel, and Amplitude as tools that generate these reports automatically; a database export into a spreadsheet also works.

What actually moves retention?

Four levers, with the first two producing the greatest results, according to the guide:

1. Focus on new users. First impressions are hard to reverse. Tactics include onboarding experiments, personalized sales, faster value hooks, and rapid feedback loops.
2. Target better users. Treat each acquisition channel — Google, Facebook, organic — as its own mini-cohort. Some channels attract ideal customers far more effectively; double down on those.
3. Kill low-value features. Overbuilding is, per the guide, the number one startup problem across industries. Expect six months to a year before you know which features to cut; use Firebase experiments, split tests, and feature flags tied to key metrics to decide.
4. Change your model or pricing. Sandbox, the author's company, shifted from transactional purchases toward a subscription offer after observing users buying repeatedly over three months — a move made to raise LTV, and one that, in the author's words, "actually worked."

What marketing tactics support retention?

Five: segmenting the customer base by usage, user type, or plan; delivering segment-relevant email; publishing how-to videos, guides, and webinars; optimizing onboarding; and using in-app messaging to mark milestones and prompt feature exploration.

The guide closes with a warning and a habit. A retention curve that falls to the floor signals a value problem, not just a pricing problem — churned customers will name pricing, but it is rarely the only reason. Retention work belongs on the daily agenda of every product and growth stakeholder, and as acquisition costs climb, the teams that instrument proxy metrics now will compound that advantage for years.