FIELD NOTES / SEPTEMBER 8, 2026 / SPECIAL REPORT

Your open rate dropped.
Did your readers disappear?

If your open rates have taken a dive lately, I understand why you’re worried. You’re still putting the work into your newsletter, and suddenly the report makes it look like your readers have lost interest.

Before you start rewriting everything or shopping for another email platform, let’s look at what that number can actually tell you.

I’ve been digging into drops in accounts I manage, and I’ve seen publishers raising similar concerns on Reddit. The pattern caught my attention around early August. I wanted to know whether something had changed at beehiiv, and what I could do about it.

This one’s longer than usual because “open rates are unreliable” isn’t much help when you’re trying to figure out what happened to yours. I’ve pulled together the numbers, beehiiv’s update, and the checks I’d start with on your next send.

CHECK WHAT’S WITHIN MY CONTROL

The Email Diagnostics Command Center can help you check your own email while you work through the bigger questions below.

FROM THE ACCOUNTS I MANAGE

A 25-point drop is worth investigating.

One established newsletter I manage had a median open rate of about 65% from mid-June through early July. By the final comparable weekly send in August, recorded opens were around 40%.

That’s roughly 25 percentage points lower, or a 38% relative decline. That’s a big enough change to take seriously.

Notice that late-July send, though. Opens had already fallen before August, then came back up a little before dropping further. I can’t honestly point to August 1 and say that’s when this account’s problem began.

These are the figures I documented for support. They weren’t all captured at the same number of days after sending, so they’re a reason to investigate, rather than a precise measurement of how much readership changed.

I’m also seeing very different results across accounts. Beehiiv’s August recap averages for six client publications ranged from roughly 15% to 68% opens. Those newsletters have different audiences, sending histories, imports and campaign types. Comparing your rate with that range won’t tell you whether you have a problem. You need to compare your newsletter with its own earlier sends.

I’ve rounded the figures and left out publication names. I also excluded a same-day result that hadn’t had time to settle. The monthly recaps cover different campaign mixes; this isn’t a controlled study.

BEEHIIV’S OWN UPDATE

What beehiiv says has changed.

Beehiiv’s Declining open rates: A diagnostic guide, updated August 25, 2026, acknowledges broader declines and outlines several mailbox-level developments:

In your own account, open the Mailbox Visibility breakdown for published posts. Look for differences between mailbox providers and groups of subscribers.

Pay attention to those dates. Beehiiv is describing several developments over time, and the guide doesn’t confirm a single platform incident beginning in August 2026.

Google’s own September 11, 2025 announcement confirms relevance-based Promotions sorting for personal Gmail accounts on mobile, with an option to return to most-recent sorting. That change dates back to 2025, so it doesn’t establish what happened to your opens this August.

ELSEWHERE IN EMAIL

Other platforms are getting these questions, too.

Mailchimp’s documentation is especially useful if your clicks fell, too. Its bot-filtering documentation explicitly warns that turning filtering on can significantly reduce reported opens and clicks. Filtering is enabled by default and applies across account reporting. Historical filtered data is available for Apple privacy opens from July 10, 2024 and email clicks from August 10, 2025. Those dates tell you how far the data goes back. They don’t establish an August 2026 rollout.

Flodesk’s July 22, 2026 update covers open-rate drops, authentication, recent sending-domain changes and testing. It’s a troubleshooting guide, without an announcement of a shared outage.

Benchmark Email’s February 17, 2026 article describes flatter rates and emphasizes clicks, visits and downstream actions. It also doesn’t identify a specific incident.

Then there’s MailerLite. Its April 15, 2026 analysis reports a 2025 median open rate of 43.46%, up from 42.35% in 2024. An annual benchmark won’t explain your August results. It does make me wary of any claim that everyone’s open rates are falling everywhere.

So yes, there’s evidence that open-rate measurement has shifted beyond beehiiv. I still can’t use those reports to explain an individual newsletter’s August drop without checking that account.

A CLICK-RATE TRAP

Your click-to-open rate can improve while clicks fall.

Beehiiv’s guide gives considerable weight to stable or rising click-to-open rate, or CTOR. I wouldn’t take that alone as reassurance that your newsletter’s doing fine. CTOR divides clicks by recorded opens, the very number you’re trying to make sense of.

Here, CTOR improves from 5% to 8%. Meanwhile, unique clickers fall 20%. You’ve got fewer people clicking, even though CTOR looks better.

When you check your reports, look at the actual counts and calculate verified unique clickers divided by delivered emails, alongside the number of people taking the action you wanted them to take. “Verified” still isn’t a promise that every event came from a human.

I ran into this while checking HMN’s own reporting: a field labeled click rate behaved like click-to-open rate. It’s worth checking what your dashboard or export is actually dividing by before you compare the percentages.

LOOK AT YOUR OWN PATTERN

Start with the pattern you actually have.

If opens fell but clickers and useful site activity held steady…
Inspect reporting filters, tracking changes and provider differences. The reporting may have changed. Check that you’re comparing the same audience over the same amount of time before assuming you haven’t lost readers.

If opens and clickers both fell…
Check delivery errors and placement, then audience mix, acquisition sources, content, offers and frequency. An email can be accepted and still be harder for readers to find. Follow up on falling click counts even if CTOR looks healthy.

If readers are telling you the newsletter’s missing…
Ask them where they found the actual production message: Primary, Promotions, Spam, or nowhere. Capture the evidence before they move it. A handful of screenshots can reveal a lead, but cannot estimate your whole list’s inbox placement.

If total traffic looks fine, go one level deeper. A busy day from search or social can conceal a decline in email traffic. Compare email-attributed sessions and your primary conversions using consistent campaign tagging. Site analytics and email clicks won’t match perfectly, but you can check whether they’re moving in the same direction.

You may find more than one of these happening in the same newsletter.

BEFORE YOUR NEXT SEND

Give yourself something useful to compare.

1. Save the evidence. Export three or four comparable sends before and after the suspected break. Add older sends for context. Compare results at the same point after sending, such as seven days, wherever you have that history. If you don’t, make a note of when you captured each result.

2. Separate the campaign types. Weekly editorial, promotions, resends and re-engagement messages answer different questions. Record imports, audience exclusions, list cleanup, frequency changes and sending-domain changes beside the metrics.

3. Split by mailbox provider. Compare delivered counts, unique opens, verified unique clickers and delivery errors within each provider. A Gmail address can be read through Apple Mail, so address domain and reading app are not interchangeable.

4. Verify the technical basics. Check authentication on received production messages and review available Google Postmaster and sender-compliance information. Google recommends keeping its reported spam rate below 0.1% and avoiding 0.3% or higher. Use Google’s own metric here; your platform’s complaint number isn’t necessarily measuring the same thing.

5. Check what changed in the sending path. In four received-message samples from the affected client, SPF, DKIM and DMARC passed, while the visible sending identity varied. I’ve asked support about that variation. Passing authentication on those four messages doesn’t settle the question of where everyone else’s email landed.

Beehiiv’s Smart Warming documentation describes gradual allocation across shared and branded sending. Progress adapts to performance, so its typical six-to-eight-week guidance for frequent senders isn’t a deadline that applies to every account. Ask for the actual allocation and any deferrals on the campaigns in question.

WAITING ON SUPPORT

Ask for evidence and a next-update time.

I’m still waiting on a technical answer in one of these cases. Support responded the day after I escalated it and passed it to Deliverability. Nearly a week later, I was following up again.

They’ve acknowledged the issue. What I’m waiting for is an explanation I can use to decide what to do next. If you’re in the same position, you know how frustrating that gap can be.

Send support enough detail that they can investigate the actual campaigns. Here’s what I’d include:

Then ask four specific questions:

01. Were there tracking or reporting changes affecting these campaigns?
02. What sending domains, IP pools and warming allocations handled them?
03. Were there provider-specific deferrals, rejections or reputation issues?
04. Who owns the technical investigation, and when should I expect the next update?

That’s my experience with this case. I can’t tell you whether it reflects a wider support backlog.

WHILE YOU WAIT

What you can do in the meantime.

Make the web edition easy to find. Share its existing URL through the channels where your readers already follow you. It gives people another way to read while you investigate delivery.

Let readers express a real preference. If they want the newsletter in Primary, Google explains how to move it between categories so Gmail can learn their preference. If a wanted message is in Spam, they can mark it not spam. Neither action guarantees placement for everyone else.

Check clipping on the actual email.Beehiiv documents Gmail’s roughly 102 KB clipping threshold and places its open-tracking pixel at the top. Clipping can hide your lower content and CTA, but you cannot automatically blame it for missing beehiiv opens. Email HTML weight is different from the file sizes of externally hosted images.

If layout is a suspect, test layout. Compare the normal design with a simpler single-column version containing the same copy, images, links and offer. Hold subject, preview, sender, timing and audience eligibility constant. Randomize within mailbox-provider groups if your tooling and list size support it.

Judge the test on verified unique clickers per delivered email after the same seven-day window, with complaints,