The only metric that matters
DAU, MAU, retention curves, bounce rate, session duration. The industry tracks dozens of metrics. For a reading product, one of them actually predicts the future.
If you open the analytics dashboard of any consumer product, you'll see a wall of metrics. Daily active users. Monthly active users. Session length. Pages per session. Bounce rate. Time on site. Return visit rate. Funnel conversion. Stickiness (DAU/MAU ratio). Net Promoter Score. Cohort retention curves. L1 to L28 retention. It's legible, it's measurable, and most of it is noise.
For a reading product, one metric actually predicts the future: week-2 retention. Did the people who visited on day zero come back at least once in the week after their first visit plus 7 days? Everything else is a leading indicator, a proxy, or a vanity number. Week-2 retention is the ground truth of whether your product is worth coming back to.
The reason is that reading is episodic. Users show up when they want something interesting. Day-1 retention is a measure of curiosity about your product. Day-7 retention is a measure of whether they remembered you exist. Day-14 retention is a measure of whether your product is good enough that they formed a habit. Curiosity and memory are cheap to create — a good landing page creates curiosity, a push notification creates memory. Habit is expensive. Habit is what separates products that compound from products that churn.
txtfeed tracks week-2 retention as the single most important number and treats almost everything else as supporting metrics. Visits this week? Useful, but not predictive. Pro conversions? Important for revenue, but not for growth. Average session length? A vanity number — longer sessions aren't always better (they might mean the user can't find what they wanted).
The discipline of picking one metric is harder than it looks. Every dashboard tempts you to optimize for the metric that's currently going up, because going-up metrics feel like progress. The trick is to know which metric is load-bearing — the one where a small move actually changes the product's trajectory — and ignore the rest most of the time.
For a reading product, the load-bearing metric is week-2 retention. Pick yours, stare at it, ignore almost everything else.
Found this useful? Share it.
See it for yourself. No signup required.
Open txtfeed