Newsletter Analytics That Drive Real Decisions

Most newsletter analytics are vanity metrics. Open rate, click rate, shiny dashboard, no action taken. After eight months of running a real newsletter at real volume, here are the few metrics that actually change what I ship next week, and the ones I stopped tracking.

The Three Metrics I Act On

Unsubscribe rate per edition. The single best signal of whether a specific issue resonated. A spike on one issue means something about that issue turned readers off. I read the comments, I check the topic, I adjust.

Growth net of unsubscribes. Subscribes minus unsubscribes minus bounces. Gross subscriber count goes up on vanity. Net growth tells me whether the product is healthy.

Reply rate and thread depth. Not 'did they open'. Did they reply. Did they start a conversation. This is the metric that correlates with conversion to paid client work. Opens do not. I know because I tracked both for three months.

Metrics I Stopped Tracking

Open rate by itself. Apple Mail privacy protections made it unreliable. The number is directionally meaningful at best. I still record it; I do not decide on it.

Click-through rate on every link. The top link is usually the lead story, which is obvious without a click tracker. Heavy tracking added no insight for my single-link-per-issue format.

Time-in-email. The providers that offer it are guessing. The number was noise.

The Decision-Making Loop

Every Monday:
  - Check unsubscribe rate on Friday's issue
  - Check net growth for the week
  - Read replies, categorize themes
  - Decide: repeat, adjust, or abandon the current series

Thirty minutes a week. One concrete decision per week. That is the operating cadence that makes analytics worth the effort.

What to Instrument

The minimum viable instrumentation: SES events for deliverability, one pixel or privacy-friendly alternative for open signals, redirect links for clicks, reply-to address that actually goes to my inbox. Four pieces. Nothing fancy. The Mailchimp benchmarks report is the best external reference for what normal looks like in your category.

The Hard-Won Rule

If a metric does not change next week's content, do not track it. A dashboard full of untouched numbers is not data; it is noise. Track what moves your decisions, cut everything else.

Newsletters are a conversation, not a broadcast. Metrics should reflect that shape.

The One-Year Look-Back

The most useful newsletter review I do is annual: which five issues drove the most replies, the most new subscribers, the most conversations that turned into paid work. The pattern is always more specific and more opinionated than I expected. Annual reviews calibrate the editorial direction better than any weekly number. If you skip everything else in this article, schedule the year-end review. It is where the data actually changes what you write next year.