What Is a Good Newsletter Open Rate in 2026?
beehiiv's State of Newsletters 2026 draws on 28 billion emails sent through its platform in 2025. It reports an average newsletter open rate of 41.24%, up from 37.98% the year before. That is the figure people keep repeating. A little farther down is the figure that gets far less attention: click-through dropped from 4.74% to 3.23% across those same emails.
In one dataset and one year, opens reached a high while clicks sank. A benchmark that praises the first figure while ignoring the second gives you a false picture. This page gives you the useful version: the real 2026 benchmarks, why some of your opens come from machines, and the two numbers that deserve more attention.
One warning before the benchmarks. The value of any benchmark depends on the sample behind it. beehiiv's numbers come from creator newsletters sent through its own platform and reported by the platform itself. The sample leans more toward independent writers than B2B corporate programmes. Even so, it is the largest public newsletter dataset of the year, and the direction of the data matches what mailbox providers say themselves. Both points matter.
What is a good newsletter open rate in 2026?
Anything above 41.24% beats the average, but every open rate you look at contains inflation. beehiiv's State of Newsletters 2026 studied 28 billion emails sent in 2025 and found a platform-wide average open rate of 41.24%, compared with 37.98% in 2024. Delivery rose from 98.31% to 98.90%. Spam complaints dropped by half, from 0.04% to 0.02%.
| Metric | 2024 | 2025 |
|---|---|---|
| Delivery rate | 98.31% | 98.90% |
| Open rate | 37.98% | 41.24% |
| Click-through rate | 4.74% | 3.23% |
| Spam complaint rate | 0.04% | 0.02% |
Source: beehiiv, State of Newsletters 2026, platform analysis of 28 billion emails sent in 2025.
Email delivery is cleaner than before. More messages are being recorded as opened. Fewer readers are clicking. Put those facts together and the benchmark question becomes less useful. The better question is simple: how much of my open rate came from an actual person?
A good newsletter open rate in 2026 is any rate you have adjusted for automatic machine opens and can compare with your own last quarter, because the raw figure combines human readers and software activity.
Why can you not trust your open rate?
Apple's Mail Privacy Protection can load email images without the subscriber doing anything, and that load can count as an open even when nobody reads the message. Apple released Mail Privacy Protection with iOS 15 in September 2021. When a subscriber enables it, Apple's servers pull remote content in the background. That action fires the tracking pixel used to count opens.
beehiiv says in its own report that MPP has made traditional open rates less reliable. Think about the incentive here. A newsletter platform benefits when writers believe their audiences are highly engaged. Large open rates support that story. Yet the same platform published its record open figure and warned readers that the metric has limits. When the company with the strongest reason to make the number look good tells you to treat it carefully, pay attention.
The move from 37.98% to 41.24% in 2025 is a real change in the measured figure. The problem is what sits inside that figure. It combines human attention with automatic image loading, and the dashboard does not give you a clean split between the two.
Did Gmail make open rates less reliable too?
Yes, and 2026 is when the change arrived. Google said on 8 January 2026 that Gmail, which has 3 billion users, is moving into what it calls the Gemini era. AI Overviews now summarise email conversations for every user at no cost. A new AI Inbox, expanding past its first testers during the year, removes what Google calls clutter and pushes messages it sees as important toward the top.
That creates two problems for your dashboard. First, software can fetch and process an email so it can summarise it. That adds another machine layer to email activity, now on the world's largest email platform instead of only Apple's. Second, and more important, some subscribers can now consume a three-line machine summary of your issue without reading the full issue. That summary does not show your design, carry your full voice, or include every middle section you worked on. It pulls out the main point. If your issue has no clear point to pull out, the summary can expose that too.
Google's announcement also makes the case for the metric this article eventually recommends. Google says its AI Inbox uses relationship signals when deciding what deserves attention. Those signals include who a user emails often, who sits in their contacts, and what relationships Gmail can infer from message content. A newsletter that earns replies creates those signals every week. Silent opens do not provide the same evidence.
What happened to newsletter click rates in 2025?
Across those same 28 billion emails, click-through dropped from 4.74% in 2024 to 3.23% in 2025. That is a fall of roughly a third in one year. beehiiv points to two causes: stronger competition in the inbox and more readers getting what they need inside the email instead of clicking through to a browser.
Both explanations make sense, and neither gives marketers much comfort. More inbox competition means your issue lands beside other newsletters written by people fighting as hard as you for attention. More in-email reading means someone can get genuine value from your newsletter without leaving a click for your tracker. So clicks can now miss some real attention while open rates count attention that never happened.
That split is the most useful fact in the 2026 numbers. Opens climbed with help from machine behaviour. Clicks dropped with human behaviour. A dashboard that depends mainly on those two metrics ends up telling you two different stories.
Which metrics should replace the open rate?
Replies and clicks on one tracked link. Both require a choice connected to a real reader. A mail app can load images in the background and create an open. It cannot type a reply for your reader, and it does not choose your single link.
Replies are stronger. Every reply gives you a named person who decided to speak to you in a thread you can come back to. In a B2B newsletter, that reply carries more value than a decorative engagement metric. It can sit at the top of a pipeline. After Gmail's January announcement, it can also help as a relationship signal because replies give an inbox AI clear evidence that two people interact.
A one-link click gives you a cleaner form of click-through. Use one link in the issue, one offer and one URL, then track that link by itself. With ten links, your click rate starts telling you about page design and choice. With one link, it tells you whether the argument moved somebody to act.
Track unsubscribes by issue too, without alarm at every move. A sharp jump after one issue tells you something about that issue. A slow and steady flow can mean weak-fit subscribers are leaving, which improves the quality of the other numbers you measure.
| Metric | What a real reader must do to move it | What can move it without a reader |
|---|---|---|
| Open | Nothing. It can fire on its own | Apple MPP image loads, Gmail processing |
| Reply | Write to you | Nothing |
| One-link click | Choose your one offer | Bots, a known limitation, kept small by using one link |
| Unsubscribe | Decide to leave | List-cleaning tools |
Should you stop tracking opens at all?
No. Keep tracking them, but move them down the pecking order. Open rate still handles two useful jobs that other metrics do poorly. Both remain useful despite machine opens because you are watching movement over time rather than trusting the raw level.
Job one is spotting deliverability trouble. If your open rate suddenly collapses on a single send, a technical problem may have appeared. It could be domain reputation, a clipped subject or a provider change. The total rate still contains machine activity, but a sudden shift can tell you that something broke. Reply volumes are usually too small to flag that kind of problem quickly.
Job two is comparing segments inside your own list. Machine opens should affect your segments in roughly similar ways. If one segment opens at half your normal average, that gap can still tell you something useful about interest, even though both rates carry inflation. That comparison helps with a decision that matters more after Gmail's January change: deciding when a segment has become quiet enough that repeated sending may teach the filter to care less about you.
Open rate should no longer be the hero metric in your report. Whether an issue worked should come from what people did: replied, clicked the one link, unsubscribed or forwarded. Opens belong on the plumbing side of the dashboard beside delivery rate. Their main question is whether something broke. Plumbing numbers matter when they fail. Most weeks, they do not deserve the spotlight.
How do you adjust an open rate for machine opens?
There is no clean formula, and accepting that fact is part of the adjustment. Standard dashboards do not separate Apple's background fetches from human opens, and Gmail's new processing creates another machine reader that dashboards do not itemise. You can still change how you use the figure. Three simple moves do the job.
Read your open rate as a ceiling instead of a reader count. "At most 41% of the list saw this" is defensible. Saying that 41% of the list read the issue goes beyond what the data can prove, and decisions based on that claim carry the same error.
Compare open rates only with your own past results, using the same list and the same platform. The machine share should stay roughly steady from month to month, which means the trend can still tell you something even when the raw level cannot. Beating a benchmark from another platform tells you very little. Falling twelve points on your own list since March tells you a lot.
Base important choices on human metrics. Use replies and one-link clicks when judging send-time tests, subject-line tests and decisions about whether to keep or cut a segment. Those actions come from people. A subject-line test based partly on opens created by Apple leaves you testing against noise.
What is a good reply rate for a B2B newsletter?
There is no public dataset for B2B newsletter reply rates, so build your own baseline. Count replies for every issue over one quarter, then try to beat that number. That is the useful method. Your own list gives you a better benchmark than somebody else's average because your audience, your niche and your ask shape the result, and no broad industry figure holds those things steady.
You only need one spreadsheet column set. Record the issue date, the number of replies, and which replies came from people who match your customer profile. After twelve issues, you have a benchmark based on your own newsletter. Nobody else gets to argue with it, including you.
Here is the full method in practice. Say your list holds 900 founders and operators, and issue one asks one question: "What did your last blog post cost you to produce, in hours?" Four people answer. You record the date, all four names, and the two who run companies you could serve. Issue two asks no question and gets one reply. Issue three ends with another one-line question and gets six. By issue twelve, the pattern is obvious. Issues with a one-line question bring in four to seven replies from your list. Issues without one bring in zero to two. Reply count has become something you can influence instead of another number you stare at.
Now compare that with the metric most dashboards lead with. Say those twelve issues averaged a 41% open rate, exactly in line with the 2026 benchmark. Across the quarter, that adds up to thousands of opens, with some unknown portion generated by Apple's servers. During the same quarter, imagine roughly forty replies, with two dozen coming from people who match your customer profile. Every one of those is a named person inside an active conversation. One of these figures can lead directly into a revenue discussion. The dashboard usually gives more space to the other one.
There is an obvious objection. Forty replies looks tiny beside thousands of opens. It is tiny. But you can verify those replies. You cannot say the same about every open. After a quarter of tracking, replies also tell you something opens never can: whether changing what you write changes the result. A better question can move a reply metric. That makes it something you can control. A metric that jumps when Cupertino changes iOS behaves more like weather.
Do newsletters still work for B2B in 2026?
The channel keeps growing on measures tied to real readers. beehiiv publishers reached more than 255 million unique readers in 2025. The share of creators making revenue on the platform doubled from 15% in the first quarter of 2024 to 30% in the first quarter of 2026. A channel with no life left does not produce that kind of doubling.
A certain newsletter model has faded: writing for a send list instead of a person and judging success mainly by a number Apple damaged in 2021. The writers making money from newsletters treat each issue like a product built for an audience. That helps explain why revenue doubled even as click rates declined. More value now lives inside the email itself. Serious readers consume it there. Casual skimmers leave less useful data behind.
The B2B case is even clearer. McKinsey's 2026 Global B2B Pulse, based on nearly 4,000 decision-makers across 13 countries, found that inconsistent information across teams is the top reason buyers switch suppliers. A weekly newsletter gives a founder one of the few places where they can control the full message, from start to finish, inside the buyer's inbox, without an algorithm sitting between them. That control deserves better measurement.
How do you get more replies instead of more opens?
Ask one question in each issue and make sure someone can answer it in one line. Questions that need a paragraph are easy to delay and forget. Questions that can be answered with a number or a yes can be handled from a phone while somebody waits in a checkout queue.
The type of question has a big effect on whether people respond. Look at the difference. "What do you think about AI in marketing?" asks for a paragraph. Most people leave it sitting there. "How many of your last ten posts did a human edit before publishing?" asks for a number, and a founder can answer in six characters. "Are you measuring your newsletter on opens or replies?" needs one word and starts the same conversation the issue covered. For a services audience, try: "What is the one page on your site you would be embarrassed to send a buyer today?" It is uncomfortable, but someone can still answer it in one line. That makes a reply more likely.
Write your question after you finish the issue. A generic question pasted onto the end feels like a survey. A question that only works because of the argument the reader finished a minute ago shows that the issue was written for a real person.
Use one link in the main body and measure that link. Add more links and you split the signal while weakening the main ask.
Send from a person's address instead of a brand address. The inbox that receives replies should also be checked by a human that same day. If someone answers your question and hears nothing for a week, you teach them that the question did not matter.
Put the main point early. Gmail's AI creates its summary from the content of your issue, and human skimmers also scan from the top. If the core point appears in the first two sentences, both the summary and the skimmer can catch it. Hide it in paragraph six and both can miss it.
FAQ
What is the average email open rate in 2026? beehiiv's State of Newsletters 2026 reports an average newsletter open rate of 41.24% from 28 billion emails sent through its platform in 2025. Its data leans toward creator newsletters, and beehiiv also says Apple's Mail Privacy Protection has reduced the reliability of open rates.
Does Apple Mail inflate open rates? Yes. Mail Privacy Protection arrived with iOS 15 in September 2021 and can load email images on its own. Each load may register as an open. Your reported rate therefore combines human opens with Apple's background fetches, and standard dashboards do not separate the two.
What is a good click rate for a newsletter in 2026? beehiiv's platform average dropped to 3.23% in 2025 from 4.74% the year before. Measure your rate using one tracked link instead of combining every link in the issue. Then compare that result with your own previous quarter rather than a broad cross-industry benchmark.
How many links should a newsletter carry? Use one link that you measure, along with the legal housekeeping links in the footer. Tracking one main link makes click-through a measure of your offer. Tracking many links makes the result depend more on the layout.
The one change to make this week
Go back to your last issue and count how many replies it received. If that count is zero while the open rate says 41%, you have a much clearer idea of what that 41% gave you. Better to learn that now than after spending a quarter of budget trying to grow the wrong number.
For the next issue, add one question that can be answered in one line and send it from an address the reader can reply to. Then count the answers. After twelve issues, you will have the only newsletter benchmark built around your business rather than another company's platform.
The four places where a buyer compares your claims, and what you need to do so they match, are covered in why B2B buyers switch suppliers in 2026. The source of the opinions that give a newsletter substance is covered in B2B thought leadership strategy. If you want another view on what your newsletter currently measures, the content agency readiness quiz takes two minutes and does not ask for your email.