Email open rate
Email open rate is the percentage of delivered emails that were recorded as opened, measured by a tracking pixel that loads when the message is displayed.
Also called Open rate, Email opens.
Email open rate is the share of delivered emails that registered as opened. It has been the headline number of email marketing for as long as there have been email tools, and it is still the first thing they show, which is a problem, because for several years now it has measured something other than people reading your email.
An open is counted by a tracking pixel. A one-pixel image with a unique address is placed in the HTML of each message, and when the recipient's mail client fetches that image, the sender's server logs an open. That is the whole mechanism, and everything wrong with the number follows from it.
The formula
Open rate is unique opens divided by emails delivered, times a hundred. The bottom of the fraction is delivered mail rather than sent mail, so bounces come out first: a campaign that sent 1,000 emails, had 50 bounce and recorded 190 opens has an open rate of 20%, not 19%. Unique matters too. One person opening the same message four times is one open.
Why open rates lie
The pixel only fires when images load. Clients that block remote images by default, which includes many corporate Outlook setups, record no open for a message somebody read closely. In the other direction, since Apple introduced Mail Privacy Protection in 2021, Apple Mail fetches every image in every message through Apple's servers whether or not anyone opens it, so anyone reading on Apple Mail with the feature on registers as having opened everything. Gmail's image proxy and the security scanners that fetch links and images do the same at smaller scale. The number now contains an unknown mix of people and machines, and no arithmetic separates them.
For cold email there is a second problem that has nothing to do with accuracy. The pixel is an image in HTML, and a first email to a stranger is supposed to look like something a person typed. A plain text message with no images is what a colleague sends; the same words wrapped in HTML with a hidden image are what a marketing platform sends, and filters have been trained on that difference. Tracking opens on cold email makes the email look like the thing it is trying not to be, and can cost you deliverability to learn a number you cannot trust.
What is worth reading instead is the chain that leads to a conversation. Delivery, confirmed by the receiving server, tells you the message arrived. A reply means somebody read it and cared enough to type; a positive reply, kept separate from the polite declines, means the targeting was right as well. If those are low, the fix is the same one that would have raised opens in the old days: write to people for whom the email is plainly relevant, use a subject line that says what the email is about, and send from a domain in good standing.
Kamu does not track opens
Kamu sends plain text and puts no pixel in it, so it has no open rate to show. The reason is the one above: an invisible image is what turns a short personal email into something a spam filter recognises as marketing, and the number it would produce would not tell a founder anything a reply does not tell them better. What Kamu records instead is delivery for every send and every reply that comes back, sorted by what it says.
In practice
- 1,000 emails sent, 50 bounced, 190 unique opens recorded: a 20% open rate.
- A campaign to an audience reading on iPhones that shows opens on almost every message and two replies, where the opens were Apple's servers fetching the pixel.
- A recipient on a locked-down corporate Outlook who read the email twice and forwarded it, recorded as never opened.
Questions
What is the email open rate formula?
Unique opens divided by emails delivered, multiplied by a hundred. Delivered is sent minus bounces, so bounced messages do not drag the rate down. Tools generally count one open per recipient however many times they return to the message, and they also count an open when the tracking pixel is fetched by a machine rather than a person, which is the reason to treat the result loosely.
What is a good email open rate by industry?
The published tables vary by industry and by the vendor publishing them, and none of them mean much any more, because the pixel they depend on is fetched by Apple's servers and blocked by many corporate clients regardless of industry. Comparing your rate against a benchmark built from the same broken measurement tells you where your audience's devices sit more than how your email did. Compare reply rates instead, against your own previous sends.
How do I improve email open rates?
Improve the things an open rate used to be a proxy for. Send to people with a specific reason to care, write a subject line that describes the email in plain words, send from a warmed domain with SPF, DKIM and DMARC in place, and keep bounces near zero. Then judge the result by replies rather than by opens, since the opens will move for reasons that have nothing to do with you.
Does open tracking hurt deliverability?
It can. The pixel requires HTML, and a short message padded with markup and an invisible remote image scores differently with filters than the same words as plain text. Marketing platforms carry that cost because their newsletters are HTML anyway. For a cold email that is meant to read as a note from one person to another, it is a cost with no matching benefit.
Related terms
Last updated