How to calculate email open rate without fooling yourself

Email open rate is simple arithmetic wrapped around a surprisingly slippery event. Divide unique opens by delivered messages, multiply by 100, and you have the reported percentage. The catch is that an "open" usually means a remote image was requested. It does not always mean a person read the message, and a real reader can open an email without loading that image.

Use open rate as a weather vane, not a courtroom witness. It can show direction across comparable sends. Clicks, replies, conversions, unsubscribes, delivery failures, and inbox placement tell you what happened next.

Email open rate formula

Most email reports use successfully delivered messages as the denominator:

Email open rate = unique open signals / delivered emails x 100

Delivered emails are the messages that did not register as bounces. If your report starts with sent messages, calculate them first:

Delivered emails = sent emails - bounced emails

Suppose a newsletter has:

• 2,400 messages sent
• 120 bounces
• 570 recipients with at least one recorded open

The delivered count is 2,400 - 120 = 2,280. The reported open rate is:

570 / 2,280 x 100 = 25%

That 25% means 570 delivered messages produced a unique open signal under the reporting system's rules. It is more precise than saying 570 people carefully read the email over breakfast.

Mailchimp's current report definitions also use successful deliveries for open rate. Check your own provider's definition before comparing reports, because bot filtering and click-as-open rules can differ.

Unique opens and total opens are different

One recipient can load the same message more than once. Keep these counts separate:

• Unique opens count a recipient or delivered message once, even if the tracking image is requested repeatedly.
• Total opens count repeated requests as well.
• Open rate normally uses unique opens, not total opens.

Total opens can help diagnose repeat activity, forwarding, or automated loading, but they should not be dropped into the open-rate numerator. Otherwise one enthusiastic reader, one forwarded message, or one busy security scanner can bend the percentage out of shape.

How to combine open rates across campaigns

Do not average campaign percentages unless every campaign has the same delivered count. Weight the result by delivery volume.

Imagine two sends:

Campaign	Delivered	Unique open signals	Reported open rate
A	900	180	20%
B	100	40	40%

The simple average is (20% + 40%) / 2 = 30%, but that gives the small campaign as much influence as the campaign sent to nine times as many people.

The combined rate is:

(180 + 40) / (900 + 100) x 100 = 22%

Use that weighted calculation for a portfolio or period total. Keep the individual rates beside it, because a combined number can hide an unusually strong or weak segment.

What an email open actually measures

Open tracking usually adds a tiny remote image with a unique URL to an HTML email. When an email client requests the image, the tracking service records the request. Pixels are useful little doorbells, but they cannot see who walked through.

The signal has several boundaries:

• A client with remote images disabled can undercount a real read.
• Plain-text email cannot load an image pixel.
• Forwarding can produce requests associated with the original tracked message.
• Security software, link preview tools, and other automated systems can create activity before a person acts.
• Some providers count a tracked click as an open when the pixel itself did not load.

Mailchimp documents bot activity as a source of inflated open and click metrics. If your reporting tool offers bot filtering, record whether it is enabled and keep that setting consistent across the periods you compare.

Apple Mail Privacy Protection

Apple Mail Privacy Protection can load remote message content privately and automatically. Apple explains that senders may see a message as opened regardless of whether the person read it, while the actual viewing time, IP address, location, and device details are hidden. Read the current Apple Mail privacy explanation before treating Apple Mail opens as person-level behavior.

This does not make open rate useless. It means a change in Apple Mail audience mix or privacy activity can move the number without an equivalent change in human attention.

Gmail image handling

Gmail serves images through Google's secure proxy servers. Google says this protects device and location details, and its sender guidance is refreshingly blunt: Google does not track open rates and cannot verify the accuracy of third-party open-rate reports. See Gmail's image handling and Google's sender guidance.

Do not use an open event as proof of a person's location, device, intent, or consent. Collect only the data you need, explain its use, honor applicable privacy rules, and give recipients the choices your messages require.

What is a good email open rate?

There is no durable universal number. A password-reset message, a weekly newsletter, and a re-engagement campaign have different audiences and reasons to open. Provider benchmarks also change with industry mix, geography, bot filtering, privacy features, and list quality.

A useful benchmark is your own comparable history:

1. Group like with like: transactional with transactional, newsletters with newsletters.
2. Compare the same definition, filtering rules, and delivered denominator.
3. Separate major audience or client segments when their behavior differs.
4. Annotate subject-line tests, sender changes, list cleanup, and privacy-filter changes.
5. Read opens beside clicks, replies, conversions, complaints, and unsubscribes.

If a campaign reports 24% this month and 21% last month, first ask whether the sends, audience, and measurement rules are comparable. A three-point change is not useful evidence when the ruler changed halfway through.

Read open rate with stronger signals

Signal	What it can tell you	What it cannot prove alone
Delivery status	Whether a receiving system accepted or rejected the message	Inbox placement or reading
Inbox placement	Whether a controlled message reached inbox, spam, or another category	Individual customer engagement
Open signal	Whether tracked remote content was requested	Human attention or understanding
Click	Whether a tracked link was requested	A successful task or purchase
Reply	Whether a response reached you	Satisfaction or conversion
Product action	Whether the recipient completed the intended journey	Why they chose to act

For a password reset, the meaningful outcome is a completed reset. For an invoice reminder, it may be a payment or a reply. For a product newsletter, it may be a visit to the relevant feature followed by use. The open is a clue near the beginning of the trail.

Diagnose a sudden open-rate change

Before rewriting every subject line, check the plumbing:

• Did the delivered count or audience size change?
• Did the provider enable or change bot and Apple privacy filtering?
• Did the share of Apple Mail, Gmail, corporate gateways, or image-blocking clients move?
• Did a template update remove, duplicate, or break the pixel?
• Did the sending domain, From name, cadence, or audience source change?
• Did click, reply, conversion, unsubscribe, complaint, or placement signals move too?

A rise in opens with flat clicks and conversions may be automated loading. A fall in opens with stable conversions may be a measurement change. A fall across opens, clicks, placement, and product actions deserves a closer look at audience relevance, sender health, and the message itself.

Test a tracking pixel with MailSlurp

MailSlurp can create a tracking pixel and report whether its URL has been requested. This small TypeScript check proves the mechanism without pretending the simulated request came from a human reader:

import { MailSlurp } from 'mailslurp-client'

async function verifyTrackingPixel() {
  const apiKey = process.env.MAILSLURP_API_KEY
  if (!apiKey) throw new Error('Set MAILSLURP_API_KEY')

  const mailslurp = new MailSlurp({ apiKey })
  const pixel = await mailslurp.trackingController.createTrackingPixel({
    createTrackingPixelOptions: { name: 'Open-rate test' },
  })

  const before = await mailslurp.trackingController.getTrackingPixel({
    id: pixel.id,
  })
  if (before.seen) throw new Error('A new pixel should be unseen')

  const response = await fetch(pixel.url)
  if (!response.ok) throw new Error(`Pixel returned ${response.status}`)

  const after = await mailslurp.trackingController.getTrackingPixel({
    id: pixel.id,
  })
  if (!after.seen) throw new Error('Expected the pixel request to be recorded')
}

verifyTrackingPixel().catch((error) => {
  console.error(error)
  process.exitCode = 1
})

When sending an HTML message with MailSlurp, set addTrackingPixel: true in the send options or add the pixel URL to the intended template. The email tracking pixel guide covers sending and webhook events, while the open tracking product page gives the direct workflow.

Test more than the event flag. Send a uniquely named message to a controlled inbox, inspect the delivered HTML, load it in the clients your readers use, and keep click and reply events separate. If an open webhook can be retried, make the receiver idempotent so one pixel request does not trigger the same follow-up twice.

Check the message people actually receive

A perfect tracking calculation cannot rescue an unreadable email. Before a customer send:

1. Deliver the real template to a controlled MailSlurp inbox.
2. Confirm the sender, subject, links, fallback text, and unsubscribe path.
3. Use device previews to inspect mobile, desktop, light, and dark rendering.
4. Run an inbox placement test when sender or campaign conditions changed.
5. Complete the customer action, not merely the pixel request.

That last step keeps the measurement honest. A reset email should reset an account. A one-time code should be accepted. A campaign link should reach a useful, working page.

Common questions

Should bounced emails be included in the denominator?

Not when you are calculating the usual delivered-message open rate. Subtract bounces first, then divide unique open signals by successful deliveries. Label a sent-message denominator clearly if your system deliberately uses one.

Can I calculate open rate from total opens?

No. Total opens include repeated activity. Use unique opens for the usual rate and report total opens separately.

Is click-to-open rate more reliable?

Click-to-open rate divides unique clickers by unique openers. Clicks can be a stronger action signal, but the denominator still inherits open-tracking noise and security systems can inspect links. Read CTOR beside delivered-message click rate and completed customer actions.

Can I tell whether one person read an email?

A pixel request can show that remote content was requested under a tracked identifier. Privacy loading, proxies, forwarding, blocked images, and automation mean it should not be treated as proof that a particular person read or understood the message.

Keep the percentage in its place

Calculate open rate with unique open signals over delivered messages. Weight multiple campaigns by their delivered volume. Then treat the result as one imperfect signal in a larger story.

MailSlurp helps you connect that signal to the useful evidence around it: the message arrived, the template rendered, the link worked, the inbox placement was healthy, and the customer could finish what the email asked them to do. That is a much better ending than admiring a percentage in isolation.