DAU/MAU is an important metric to measure engagement, but here's where it fails
DAU/MAU became the gold standard because Facebook's was always >50%, but most billion-dollar companies have terrible DAU/MAU—and that's fine, because fighting your product's natural usage cadence is futile.
Read Original Summary used for search
TLDR
• Facebook popularized DAU/MAU (>50% is "world class"), but this metric only works for daily-use products like messaging/social—most valuable products are naturally episodic
• Uber's most profitable rides are airports/special occasions, LinkedIn is only used by recruiters/job seekers in spurts, Airbnb/travel products used ~2x/year—all multi-billion dollar companies with low DAU/MAU
• Flurry data shows app categories cluster predictably: social games = high frequency/low retention, weather = low frequency/high retention, communication = both high
• Trying to boost DAU/MAU with notifications typically backfires—increases MAU more than DAU, actually lowering the ratio; products rarely jump from 10% to 40% through effort alone
• Better metrics for low-frequency products: % hardcore daily users, correlation between usage and network size/content creation, monetization per transaction
In Detail
Andrew Chen challenges the obsession with DAU/MAU (daily active users / monthly active users), arguing it's only relevant for high-frequency, ad-supported products. Facebook popularized the metric because theirs was always exceptional—over 50% even in 2004 with just 70k users. But this created a false standard: most successful products have naturally episodic usage patterns. Uber's most profitable rides are infrequent (airports, Black Car, business travel), LinkedIn is only used by recruiters and job seekers in spurts, travel products like Airbnb are used ~2x/year, and most SaaS tools (Workday, Salesforce, Google Analytics) are used 1-2x/week at most. All are multi-billion dollar companies with "terrible" DAU/MAU.
Flurry's retention vs. frequency data reveals predictable clustering by category: social games have high frequency but low retention (burn through content), weather apps have low frequency but highest retention (lifelong utility), and communication products have both. The key insight is that usage cadence is determined by product category, not effort. Attempting to increase DAU/MAU through notifications typically backfires—it grows MAU faster than DAU, actually lowering the ratio. Chen notes he's never seen a 10% DAU/MAU product become 40% through sheer effort.
For products with naturally lower frequency, Chen recommends different metrics: identify your hardcore users (what % were active every day last week?), show how usage frequency correlates with network size or content creation (proving network effects), or simply focus on monetization per transaction. The lesson is to stop fighting your product's natural cadence and find metrics that prove you're delivering value in the pattern that makes sense for your category.