Which numbers tell you whether readers are finding you?

Which numbers mean something?

The numbers that mean something record a person choosing to act. Look for these, whatever services you use:

• Net subscribers: sign-ups minus departures, month by month. A list that grows slowly and keeps its readers is a healthy one.
• Unsubscribes and spam reports per issue: the plainest verdict any single letter gets.
• Replies and comments: someone took the trouble to write back.
• Reads: if your blog’s analytics separate reads from views, reads show that someone stayed and scrolled.
• Where sign-ups came from: the page, post or event that brought each new subscriber, so you know what to repeat.
• Plays per episode: how each new podcast episode does in its first weeks, compared with earlier episodes at the same age.
• Sales: your retailer or publisher reports. A click on a buy link records interest, and only the sales report records a purchase.

Compare every number with your own earlier months. Another writer’s figures were counted another way, for another audience, and won’t tell you much about yours.

Why are open rates unreliable?

Open rates are unreliable because software can register an open with no reader involved. An open is usually counted with a tiny invisible image in each email. When the reader’s mail app fetches that image, the sending service logs an open.

In June 2021, Apple announced Mail Privacy Protection, designed to stop senders using those invisible images to learn when someone opens an email (source). With the feature on, Apple’s Mail app downloads remote content in the background, tracking images included. It does this whether or not the reader engages with the message (source). It also hides the reader’s IP address, so location figures drawn from opens are unreliable too.

So an issue can show as opened when nobody read it. Opens can also go uncounted when a reader’s app hides images, which means the figure is wrong in both directions, by an unknown amount. Don’t use it to judge a subject line or to compare one issue with another.

Watch these instead:

• Traffic on send day: give each issue one main link, then check visits to that page on the day you send.
• Clicks, if your service tracks them: a click usually needs a person, so it holds up better than an open, though no inbox figure is perfect.
• Replies: ask a real question now and then, and count the answers.
• Unsubscribes and complaints per issue: compare them across several letters.

What should you check each month?

Once a month, spend ten minutes on the same short list and write down what you find. A notebook or a simple spreadsheet is enough.

• Your list: subscribers gained, lost and net, against last month.
• Your weakest letter: the one with the most unsubscribes or complaints, and what was different about it.
• Your strongest post: the one with the most reads, comments or shares, and what it has in common with earlier strong ones.
• Where newcomers came from: the pages, posts or events behind new sign-ups.
• Your listening: each recent episode’s early plays against earlier episodes at the same age.
• Your sales: any jump in your retailer or publisher reports, and what you did that week.

Then choose one thing to try next month, and only one, so you can tell whether it worked. Small numbers swing a lot from month to month, so look at the last three months together before deciding anything has changed.

Which numbers can you ignore?

You can ignore any number that can grow without anyone reading, listening or buying. These are the common ones:

• Follower counts: they can be bought or swapped, and they don’t show who sees your posts.
• Likes and impressions: a network’s tally of taps and screen appearances, counted by its own rules.
• Raw page views: a view records a page loading, including visitors who left within seconds.
• Open rates: for the reasons above.
• All-time totals: downloads or views since the beginning can only rise, so they can’t tell you whether this month went well.
• Other writers’ numbers: different readers, different counting.

Chasing these numbers backfires, because the tactics that inflate them attract people who don’t read. A giveaway that trades a prize for an email address can fill a list with people who only wanted the prize. Some will leave or mark you as spam, which hurts delivery to the readers you want. A clickbait title raises views and spends trust, and bought followers make your genuine engagement look thin beside the total.

A small audience that reads everything you send is worth more than a large one that scrolls past. Grow that one, and measure it honestly.