What makes a newsletter successful?#
A newsletter succeeds when the action it was written to cause happens often enough to be worth writing. That action differs by job, and the metric has to follow it. A sales newsletter is successful when a reader buys. A nonprofit newsletter is successful when a reader gives, attends or turns up as a volunteer. A personal newsletter is successful when a reader writes back. Only one of those three is measured by an open rate.
The reason this matters in practice is that the report you get from any email tool stops at the mailbox boundary. It counts what happened to your message: delivered, opened, clicked, unsubscribed. What happened after a click belongs to your website, your checkout or your event page, and no email tool sees it unless you put a marker on the link and read the analytics on the other side.
Is a high open rate a sign of success?#
It is the weakest signal in the report, and the vendor that publishes the most-used figures says so in its own disclaimers. The Mailchimp benchmarks page states that the accuracy of email open rates may be impacted by Apple's Mail Privacy Protection, and that this should be considered when interpreting open rate data. That feature preloads images on Apple devices, which registers an open before a human eye reaches the message.
An open also answers a question you already know. It reports curiosity, not value: nobody opens a newsletter they dislike twice, but plenty of people open one, read none of it and never open the next. If you want the deeper read on the denominator, what a good open rate is works through the arithmetic.
Keep opens for one narrow job: comparing a subject line against your own average. They are a good signal about the envelope and a bad signal about the contents.
What numbers do email providers publish?#
One page carries the figures most cited elsewhere, so it is worth reading with its conditions attached rather than as a set of targets. Mailchimp calculates averages from emails delivered through its own system, where campaign tracking was switched on and users reported an industry, and it states the figures below were last updated in December 2023.
| The figure on the Mailchimp benchmarks page | What it publishes | The conditions it attaches |
|---|---|---|
| Average open rate | Around 34.23 percent as the figure to aim for, with 40.55 percent for government as the highest industry and 27.34 percent for vitamin supplements as the lowest | Unique opens, measured across campaigns sent to at least 1,000 subscribers, and the page warns Apple's privacy feature affects accuracy |
| Average click rate | 2.66 percent described as optimal, with a range of 1 to 5 percent by industry, 1.19 percent lowest for vitamin supplements and 4.58 percent highest for government | The page defines its own CTR as total clicks divided by emails delivered |
| Average unsubscribe rate | 0.22 percent across its users, with 0.15 percent business and finance, 0.18 percent non-profits and education and 0.19 percent ecommerce | Same 1,000-subscriber floor, and data marked as last updated in December 2023 |
| Bounce figures | Soft bounce rate and hard bounce rate calculated by industry | Measured on campaigns reaching at least 1,000 subscribers, so a small list is not represented in the sample |
| Coverage | Billions of emails delivered by its system, across one-person startups and large companies | The page also states these figures may vary from the benchmarks shown inside the Mailchimp application |
Read those as a shape rather than a grade. They tell you where ordinary sits. They cannot tell you your number, because they were measured across strangers with different lists, different denominators and a privacy feature that inflates one of the three columns.
How do I measure the success of a newsletter?#
Work down a ladder of five rungs and pick the rung your goal sits on. Each rung has its own number, and each one answers a different complaint.
- Delivered: did the mailbox accept it at all. This is where bounce and complaint problems show up, and it gates every rung below.
- Opened: did the envelope interest anybody. The noisiest rung, and the one a privacy feature inflates.
- Clicked: did the contents move somebody to act inside the message.
- Acted: did the click turn into the thing you wanted, meaning a purchase, a booking, a donation, a reply or a registration.
- Stayed: are the same readers still there in three months, and is the list still large enough to matter.
Two of those rungs are the honest answer for most newsletters, and neither is in the email tool by default. The fourth needs a tracked link and a look at your site analytics. The fifth needs you to note the count today and compare it later, which is the measurement almost nobody does and the one that reveals whether the work is compounding.
Why do two tools disagree about my rates?#
Same words, different maths, and it happens inside our own app. The Mailchimp benchmarks page describes its click-through rate as total clicks divided by the total number of emails delivered. Our campaign report shows the clicks tile as a percentage of openers instead, which produces a different figure for the same send, and neither of us is wrong. It is why a number carried from one tool into a slide should arrive with its denominator.
| The metric | Definition on the Mailchimp benchmarks page | What that means when you compare tools |
|---|---|---|
| Open rate | Calculated as average unique open rates | A total-opens figure on another tool sits higher on the same campaign, and the gap grows with repeat readers |
| Click rate | Total clicks divided by emails delivered | A clicks-over-opens figure on the same campaign reads far higher, so the two are not the same row |
| Delivered | The denominator used for its click rate | A campaign with heavy bounces shows a lower click rate on that basis than on an opens basis |
| Unsubscribe rate | Average 0.22 percent across its users | Small campaigns are excluded by its 1,000-subscriber floor, so a small list has no true peer figure |
| Bounces | Soft and hard bounce rates published by industry | A hard-bounce problem pushes every other rate on the report down at once |
One rule settles it: quote the count and the denominator together. Say 61 clicks out of 1,900 delivered, and a stranger can rebuild any rate you like from it. Say 3.2 percent and you have started an argument you cannot finish.
What are the right KPIs for a newsletter?#
Choose two: one that measures this send and one that measures the list. A send KPI tells you whether today's work landed. A list KPI tells you whether the whole project is getting stronger or quietly thinner.
- Sales newsletter: actions completed per send, and revenue per send, with click rate as the early warning.
- Community or nonprofit newsletter: replies, event signups and donations per send, with unsubscribe rate as the caution light.
- Paid or media newsletter: signups from your own site, conversion from free to paid, and how many subscribers remain active after a fixed window you set yourself.
- Personal or expert newsletter: replies per send and forwards per send, since both mean a reader did something no report counts.
Two to three KPIs in a working document is as far as most small senders can carry. Ten numbers and you will find the one that flatters you. That is my reading of how these sets behave, not a published standard.
How do I know if my newsletter is growing?#
The formula is plain and it is the part a tool never writes for you. Change in your list equals new addresses added, minus unsubscribes, minus addresses that die on their own when a mailbox is abandoned or a person leaves a job. A list that shows the same total every month is not stable, because the third term is running underneath it.
Published decay figures exist on vendor pages, and none of them are auditable on your list, because nobody can see which of your addresses were ever reachable. The usable version is to record three numbers once a month, the same three for as long as the newsletter exists: added, removed by choice, removed by failure. Two of those three are already in your report, and the third comes from the bounce log.
What does a good unsubscribe rate look like?#
Mailchimp publishes 0.22 percent across its users, and the industry split sits between 0.15 and 0.19 percent on the four sectors it names. Treat those as shape rather than a grade, and read the exit as a message rather than a failure: an unsubscribe is a reader who told you something instead of reporting you, and the report button is the alternative they did not take.
The full arithmetic, both published benchmarks and what each shape of jump means, is at what a good unsubscribe rate is.
Should I measure clicks or conversions?#
Both, in that order, because they answer different questions about the same send. A click says the message did its job at the moment of reading. A conversion says the offer was real enough to complete. A campaign with a strong click rate and no conversions has a copywriting win and a product problem, and the email report alone will keep telling you the good news.
The gap is worth tagging for: put a campaign marker on the link, read the destination in your site analytics and keep the spelling the same every send. Should I tag links to see where clicks went covers how the tagging works and where it breaks.
How often should a newsletter go out?#
Consistency beats frequency, and the reason is a promise. A reader who learns that your thing arrives on Tuesday morning has a place for it in the week. A reader who cannot predict when it lands has one option left, which is the archive or the unsubscribe. Sending more often does not create the habit and sending too rarely dissolves it.
The working rule from our own data, measured on the reports rather than on a theory: engagement holds while readers can predict you, and the first number to move when they cannot is the unsubscribe rate. How often should you send a newsletter takes the question apart properly.
What makes readers keep a newsletter?#
Three things show up in every newsletter somebody describes as worth keeping: it is useful at a specific job, it is recognisable before it is opened, and it can be scanned in the time a person has on a Tuesday morning. The first is content. The second is a name, a sender and a subject-line habit. The third is formatting, and it is the one most writers damage by writing as though every reader will arrive with twenty free minutes.
The first email a new reader receives is where all three get set, which is why what a welcome email should do is the cheapest success lever available to a small sender, and the least used.
Do subject lines decide whether a newsletter works?#
They decide the open, and the open decides the chance your work gets read. That is real influence over the first rung of the ladder and none over the fourth. A subject line that overstates the contents buys an open and spends reader trust, and the next send is measured against the discount it earned.
The craft side, with the patterns that hold up and the ones that cost you, is at email subject lines that get opened.
What is the smallest test worth running?#
One variable, one audience, one decision made before you start. Split the send, hold everything else still, and let the winner be picked by the count rather than by taste. That is what a subject-line test is for, and our own campaign report carries the result with a named winner once the test concludes.
- Pick the one number that decides, before the test starts, rather than after you see which way the numbers fell.
- Change one thing only. Two changes and the result cannot be read at all.
- Split the audience in halves rather than slices so each version reaches enough people to produce a difference worth naming.
- Write down what you learned in one line, because a test nobody recorded has to be run again next quarter.
Skip testing for a list of a few hundred, where one reader moves the percentage by half a point. There, use counts, and read several sends together rather than one.
Which numbers should I ignore?#
- Total opens, which counts a reader looking twice and calls it two people.
- Subscriber count on its own, which says nothing about how many of them still read you.
- A single campaign against an industry average, when the average was measured on a denominator you cannot see.
- Any figure a tool shows without saying what it was divided by.
How do I set a target without a benchmark?#
Take your own baseline. Send three times to the same audience with nothing unusual in them, note the middle result for the one number you decided to improve, and set the target as a change in that single figure over the next quarter. Not five figures at once, since two moving numbers cannot be explained when one of them turns out to be a bounce problem.
This is a working reading rather than an industry standard, and the honest reason is that no published process exists for a sender whose list falls under a tool's reporting floor. Your own three sends are the only comparison where the audience, the offer and the denominator all match.
Does deliverability belong in a success report?#
It belongs at the top of it, because it is the precondition rather than one metric among many. Mail that lands in a spam folder is opened far less often, and a report that blames the subject line for that number is measuring the wrong thing. Google publishes a spam rate limit for bulk senders of below 0.3 percent as measured in its Postmaster Tools, which is a deliverability number rather than an engagement one, and it caps the rungs below it.
Check placement before you judge content: why your email goes to spam, and what sender reputation is.
Our campaign report labels every number with the denominator it was divided by, separates delivered from addressed, keeps unique opens apart from total views, and shows an A/B subject test with a named winner once it concludes. It sends from your own domain, and the free plan is the whole app up to 5,000 emails a month.
Send a newsletter with every number labelled