How do you know whether a newsletter is working?
Six numbers, one question: did anyone act?
Why can't open rate answer the question anymore?
Open tracking works by hiding a tiny image in the email. When the reader's mail app loads the image, the platform records an open. Mailchimp's own help page explains the two ways this fails. If the reader or their email client blocks images, no open is recorded, even when the email was read. And when something loads the image without a person reading, an open is recorded anyway.
Apple Mail is the big case of the second failure. With Protect Mail Activity on, Apple downloads remote content in the background by default, regardless of whether the reader engages with the email, and routes it through relays so the sender cannot see the reader's IP address. Mailchimp names Apple's Mail Privacy Protection, along with spam filters and bots that follow links, as activity that can falsely inflate opens. So your open rate is a mix of real reads, silent machine loads and missing reads from image blockers. It can still flag a delivery problem or a subject line that flopped. It cannot tell you whether anyone found the issue useful. For how to read the number against third-party averages, see our open rate benchmarks.
Which metrics actually tell you whether a newsletter is working?
Use the table below as the framework. Each row says what the metric measures, what breaks it, and what to do about it. The order runs from the weakest signal to the strongest.
| Metric | What it measures | What breaks it | What to do |
|---|---|---|---|
| Open rate | Whether the tracking image loaded. A rough check that the email arrived and the subject line earned a look. | Apple Mail loads it without a human. Image blocking and plain text hide real reads. Security scanners inflate it. | Watch for sudden drops only. Never report it as engagement. Segment Apple Mail readers out if your platform allows it. |
| Click rate | Clicks divided by emails delivered. Shows how many readers wanted more than the email gave them. | An issue with no link scores zero. Bots that scan links add fake clicks. One link at the top beats five at the bottom. | Put one clear link per issue. Compare against your own last six issues before any benchmark. |
| Click-to-open rate | Clicks divided by opens. Shows whether the content matched the subject line's promise. | Inherits every open rate problem. An inflated open count drives this number down for no real reason. | Use only to compare two issues sent in the same week to the same list. Do not track it month to month. |
| Replies | How many people wrote back. The clearest sign that a human read and cared. | A no-reply sending address kills it. Automated out-of-office replies pad it. | Send from a real inbox. Ask one question per issue. Count real replies by hand; it is a small number. |
| Unsubscribe rate | Unsubscribes divided by delivered. Shows how many readers decided the newsletter is not for them. | A hidden unsubscribe link pushes people to the spam button instead. A promotion to a cold list causes a spike that is not about the content. | Keep a visible unsubscribe link and support one-click unsubscribe. Investigate a jump after a specific change. |
| Spam complaint rate | Share of delivered mail that readers mark as spam. Gmail reports it in Postmaster Tools. | Sending to people who never opted in. Long gaps between sends, so readers forget you. Only Gmail readers are counted. | Stay under 0.10% and never reach 0.30%, per Google. Above that, pause promotions and clean the list. |
| Net list growth | New subscribers minus unsubscribes, bounces and complaints. Shows whether the audience is actually getting bigger. | Counting signups alone hides churn. A giveaway adds people who leave within two issues. | Report the net number every month. Note where the new subscribers came from. |
| Business signal | Meetings, replies or deals where someone mentions the newsletter. The reason it exists. | Nobody writes it down. Sales never asks how the lead heard of you. | Add one field to the CRM: mentioned the newsletter, yes or no. Review it monthly with sales. |
Two definitions, since the table uses them. Click-to-open rate is clicks divided by opens, so it tries to measure the content on its own. Net list growth is signups minus everyone who left, so it measures the list, not the signup form.
What is a good number?
The honest answer is: better than your own last six issues. Vendor benchmarks are averages across every industry and every list size. Mailchimp's benchmark, for example, only counts campaigns sent to at least 1,000 subscribers, and Mailchimp itself notes that Apple's privacy changes may affect the open rate figures. A B2B list of 1,400 people in one niche will not behave like that average, and should not try to.
The two exceptions are the thresholds Google publishes for its own inbox. Keep the spam rate shown in Postmaster Tools below 0.10%. Never reach 0.30%; Google says bulk senders above that line are not eligible for help getting their mail delivered. Those are limits, not goals. For a B2B list, the spam rate should sit near zero most months.
What is the business signal, and how do you catch it?
The business signal is any moment where the newsletter shows up in a sales conversation. A prospect who replies to an issue and asks for a call. A lead who tells the sales rep they have read the newsletter for a year. A customer who forwards an issue to a colleague who then books a demo. None of these appear in your email platform's dashboard, which is why most teams never count them.
Catching it takes two habits. First, send from a real address and read the replies yourself, or have one person own that inbox. Second, add one yes-or-no field to the CRM, "mentioned the newsletter", and ask the sales team to fill it in on every first call. The number will be small. Five mentions in a quarter from a list of 2,000 is a strong result, because each one is a real conversation, not a loaded image. If you are still setting the newsletter up, our guide to running a B2B newsletter covers the reply-first setup.
How do you review a newsletter each month?
Block 30 minutes on the first working day of the month. Pull the numbers for every issue sent last month and answer these five questions in order. Write the answers down, one line each, so next month has a baseline.
- Did the mail arrive? Check bounces, the spam rate in Postmaster Tools, and whether open rate fell off a cliff on any single issue. A cliff means a delivery problem, not a content problem. Five minutes.
- Did anyone act? Compare click rate and reply count per issue against the previous month. Note which issue got the most clicks and what its one link was. Ten minutes.
- Did anyone leave, and why? Look at unsubscribes and complaints per issue. A flat line is fine. A spike on one issue means read that issue again and ask what changed. Five minutes.
- Is the list bigger than it was? Net growth: signups minus everyone who left. Note the top signup source. If growth is negative two months running, the growth plan needs work before the content does. Five minutes.
- Did the newsletter show up in the business? Count CRM records with the "mentioned the newsletter" field ticked, plus any reply that turned into a call. Five minutes.
Keep the output to one page: a short table with a row per issue and a column per question, plus three lines of notes. The point is a record you can read in a year, not a dashboard you look at once.
What do you change first when a number is off?
Change one thing at a time, and change the cheapest thing first. The pairing below is the order Sawa works through. It is a planning rule, not a guarantee; a small list will show noise that looks like a trend.
Open rate dropped on every issue at once
Suspect delivery, not content. Check authentication, bounces and the Postmaster Tools spam rate before touching a subject line.
Opens are steady, clicks are falling
The subject line is doing its job and the body is not. Cut the issue to one idea and one link. Our subject line guide covers the opposite case, where the body is strong and the subject line undersells it.
Unsubscribes or complaints spiked on one issue
Re-read that issue. Common causes: a sales pitch to a list that signed up for insight, a send to a segment that never opted in, or a long gap since the last issue. Fix the cause, then say so in the next issue.
Nothing is wrong, nothing is growing
The content is fine and the list is stale. Move effort to signup sources: a lead magnet, a LinkedIn-to-email routine, a partner swap. Also check that the schedule matches what you promised at signup; the cadence framework explains why a missed promise shows up as silence rather than complaints.
No business signal after six months
Check the two habits above first. If replies arrive but never turn into calls, the issue is not asking for the next step. If replies never arrive, the content is not about the reader's problem.
How long before you can trust a trend?
On a list under a few thousand people, one issue is an anecdote. Ten clicks more or fewer is normal variation, and a single reply can swing the reply count by half. Judge the newsletter on a month of issues, and judge a change on at least two months after you made it. Compare like with like: the same list, the same day of the week, the same kind of issue. If you changed the subject line style, the send day and the length in the same month, no metric can tell you which one mattered. That restraint is the whole method. Six numbers, reviewed on a schedule, with one change at a time between reviews.
About this guide and its sources
This is Sawa's review framework for client newsletters. The metric table and the five-question routine are planning rules, not research findings. The spam rate thresholds are Google's published limits for Gmail. The Mailchimp figures are a third-party average across Mailchimp customers and are not targets for a B2B list.
- Google: Email sender guidelines
Supports the 0.10% and 0.30% spam rate thresholds, the 5,000 messages a day line, and the one-click unsubscribe requirement.
- Google: Email sender guidelines FAQ
Supports the 48-hour unsubscribe recommendation, that spam rate is calculated daily in Postmaster Tools, and that senders above 0.3% are ineligible for mitigation.
- Apple: Mail Privacy Protection and Privacy
Supports that Protect Mail Activity downloads remote content in the background by default, regardless of whether the reader engages, and routes it through two relays.
- Apple: Protect email privacy in Mail on Mac
Supports that the reader's IP address is hidden from senders and remote content is privately downloaded in the background.
- Mailchimp: About open tracking
Supports how the tracking image works, why blocked images and plain text hide opens, and why Apple MPP, spam filters and bots inflate them.
- Mailchimp: Email marketing benchmarks
Supports the 35.63% open, 2.62% click and 0.22% unsubscribe averages, the December 2023 data date, the 1,000-subscriber minimum, and the Apple caveat.
Frequently asked questions
Is your sending schedule part of the problem?
A missed cadence promise shows up as silence in the metrics. The finder suggests a starting schedule from three questions.
Need help with this? See how Sawa handles newsletters and engagement, or book an intro call.