Engagement metrics that predict churn early

# Engagement metrics that predict churn early

A customer success manager at a mid-sized SaaS company opens her dashboard on a Monday morning. Revenue looks fine. Logo count is flat. But three enterprise accounts, worth a combined $180,000 in annual recurring revenueannual recurring revenueAnnual Recurring Revenue (ARR) is the normalized, predictable revenue a subscription business expects to earn from active contracts over a single year.Voir la définition complète → (ARRARRAnnual Recurring Revenue (ARR) is the normalized, predictable revenue a subscription business expects to earn from active contracts over a single year.Voir la définition complète →, the yearly value of subscription contracts), haven't touched the product's core feature in 18 days. Six weeks later, all three cancel. The invoice never warned her. The usage data did, if anyone had been watching.

This is the core problem with churn: by the time it shows up in revenue, it's too late to act. Engagement metrics are the early warning system. This lesson covers three of the most predictive: activation rate, feature adoption depth, and DAU/MAU stickiness.

Why revenue metrics lag and engagement metrics lead

Churn rateChurn rateChurn rate is the percentage of customers or revenue lost over a period. It measures how fast a business loses its existing customer base.Voir la définition complète → and net revenue retentionnet revenue retentionNet Revenue Retention measures the percentage of recurring revenue retained and grown from existing customers over a period, including upsell and expansion, net of downgrades and churn.Voir la définition complète → (NRRNRRNet Revenue Retention measures the percentage of recurring revenue retained and grown from existing customers over a period, including upsell and expansion, net of downgrades and churn.Voir la définition complète →, the percentage of recurring revenue kept and expanded from existing customers over a period) tell you what already happened. They're lagging indicators.

Engagement metrics tell you what's about to happen. A customer who stops logging in doesn't cancel immediately. There's usually a gap of weeks or months between "disengaged" and "cancelled," and that gap is where retention teams either intervene or lose the account.

The goal isn't to track more metrics. It's to track the *right few* that reliably move before revenue does.

Activation rate: did they ever get value?

Activation rate is the percentage of new users or accounts who complete a defined set of actions that signal they've experienced the product's core value, within a set time window (commonly 7, 14, or 30 days).

The definition of "activated" varies by product. Examples:

• Slack: a team sends 2,000+ messages (an early internal benchmark reportedly used by Slack)
• A project management tool: user creates a project AND invites a teammate AND completes one task
• An analytics tool: user connects a data source AND builds one dashboard

Why it predicts churn: accounts that never activate churn at dramatically higher rates, often within the first billing cycle. Industry benchmarks (OpenView Partners' annual SaaS benchmarks report, a widely cited free resource, openviewpartners.com) suggest well-performing B2B SaaS products see 30 to 50% activation rates within the first week for self-serve signups, though this varies enormously by product complexity.

Worked calculation:

• New trial signups this month: 1,000
• Users completing the activation action set within 14 days: 320
• Activation rate = 320 / 1,000 = 32%

If your activation rate sits below 25% for a self-serve product, expect elevated early churn regardless of what your pricing or onboarding emails say.

Feature adoption depth: are they using enough to stay?

Activation asks "did they start?" Feature adoption depth asks "are they going deep enough to be dependent?"

Feature adoption depth measures how many of a product's core features (not just any feature, the ones tied to retention) an account actively uses over a rolling period, typically 30 or 60 days.

Accounts using one feature are fragile: a single workflow change or champion departure and they're gone. Accounts using three or more interconnected features are structurally harder to replace, because switching costs compound.

Example: HubSpot customers who adopt marketing, sales, and service hubs together show materially higher retention than single-hub users, a pattern common across multi-product SaaS platforms. The mechanism is simple: more embedded workflows equal higher switching cost.

How to build a simple adoption depth score:

adoption_score = count(core_features_used_by_account_in_30_days)

Tiering example:
0-1 features used  -> "at risk" tier
2-3 features used   -> "developing" tier
4+ features used    -> "sticky" tier

Segment your churn ratechurn rateChurn rate is the percentage of customers or revenue lost over a period. It measures how fast a business loses its existing customer base.Voir la définition complète → by tier. In most B2B SaaS portfolios, the "at risk" tier churns at 3 to 5x the rate of the "sticky" tier, an estimate consistent across multiple customer success benchmarking studies (as of 2024 to 2025 data).

DAU/MAU stickiness: how often do they come back?

DAU/MAU ratio (daily active users divided by monthly active users) measures stickiness: what share of your monthly user base engages on a given day.

Formula:

Stickiness ratio = DAU / MAU

Worked calculation:

• Daily active users (average across the month): 4,500
• Monthly active users: 18,000
• Stickiness = 4,500 / 18,000 = 25%

Benchmarks (estimates, vary heavily by product category):

• Consumer social apps (e.g. the category Meta's products compete in): often 50 to 60%+
• B2B productivity/collaboration tools (Slack-like): commonly cited estimates around 30 to 45%
• B2B tools used weekly or monthly by design (e.g. payroll, quarterly planning software): naturally low, 10 to 20%, and *this is not automatically a bad sign*

This is the critical nuance for a non-technical audience: stickiness benchmarks are product-dependent, not universal. A payroll tool used twice a month is healthy. A team chat tool used twice a month is dying. Always compare a stickiness ratio against its own historical trend and against comparable products, never against an arbitrary industry-wide number.

Why it predicts churn: a declining stickiness trend, say from 35% to 22% over two quarters for a daily-use product, is one of the most reliable pre-churn signals available, often visible 60 to 90 days before a cancellation request.

Building the leading indicator system

Individually, each metric is useful. Combined into a simple scoring model, they become a churn early-warning system.

A basic version, buildable in a spreadsheet or a customer data platformcustomer data platformA Customer Data Platform unifies customer data from all sources into persistent, actionable profiles that other systems can use.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 →, a tool that unifies customer data across sources):

| Signal | Weight | Trigger for "at risk" flag |

|---|---|---|

| Activation status | 25% | Never activated within 30 days of signup |

| Feature adoption depth | 35% | Dropped from "sticky" to "developing" tier or below |

| DAU/MAU trend | 40% | Stickiness ratio declined 15%+ over trailing 60 days |