Left on Read: What a Read Receipt Actually Does

Somebody has not replied to you. There are two versions of that.

In the first, you do not know whether they have seen it. In the second, a small word underneath your message says Read.

The silence is identical. What changes is what you conclude from it, and that change is measurable.

The number that explains the whole thing

Hoyle, Das, Kapadia, Lee and Vaniea surveyed 718 Facebook Messenger users in 2015 and published the results at CHI in 2017. They asked senders how they felt when a message went unanswered, splitting by whether the receipt had fired.

Reaction	Message not read	Message read, not replied to
Upset or angry	30.3%	42.8%
Felt slighted or ignored	12.7%	26.9%
Not a big deal	18.4%	9.2%
Concerned about the recipient	6.2%	1.5%

And when they asked what people assumed about the other person:

Attribution	Not read	Read, not replied to
“Is busy”	24.1%	11.7%
“Is pointedly ignoring me”	15.4%	27.9%

Read that last row again. The same non-reply is charitably explained as busyness when the receipt has not fired, and read as deliberate snubbing when it has. The receipt does not tell you anything about why someone has not replied. It changes what you assume the reason is.

That is what a read receipt does. It converts an ambiguous silence into an apparently informative one, and the information it seems to supply (they chose not to answer) is not information at all.

Its biggest measured effect is not on replying

Here is the finding that ought to be better known.

In the same study, 68.4% of recipients said they had deliberately avoided opening a message because of the “seen by” feature. Of those, the most common reason, given by 68.2%, was: “I wanted to pretend I never saw the message.”

Second most common: too busy to view it, 45.8%. Third: hadn’t responded and didn’t want the other person to know they had logged in, 41.3%.

A feature built to remove ambiguity is used, by two-thirds of recipients, to manufacture it.

The mirror image also happens. 28.2% said they had deliberately opened a message because of the receipt, 86.7% of them “to show the other person that I saw his/her message,” and 47.0% “to show I am not ignoring them.” The receipt has become a cheap substitute for a reply: an acknowledgement you can send by doing nothing but looking.

And on the sender’s side, 88.58% of receipt-aware senders said they used the information to see who had received a message but not responded.

So the feature is doing a great deal of work. Just not the work it is usually described as doing. Which brings us to the thing that is missing.

Nobody has shown that read receipts make people reply faster

This is the obvious claim, it is intuitively compelling, and as far as we could establish, it has never been tested.

We looked for an experiment or a log study measuring reply speed with receipts on versus off. There is none that we could find. The largest messaging-log study we located, 3.4 million messages across 889 chats, does not mention read receipts at all. It does tell us that around 70% of WhatsApp messages are answered within five minutes, and that reply speed is strongly reciprocal between two people, but it says nothing about receipts.

So the state of the evidence is: strong, replicated self-report evidence that receipts change how people handle opening messages, and no evidence whatsoever about whether they change replying. The one thing everyone assumes is the one thing nobody has measured.

People do not understand what the ticks mean

Two separate studies found the same thing, eight years apart, and it is a design finding rather than a user failing.

Church and de Oliveira interviewed nine WhatsApp users in Spain in 2013, a year before blue ticks existed. Seven of the nine believed the second grey tick meant the message had been read. It meant delivered. Users had invented a read receipt before the feature was built.

Then Hoyle and colleagues found that 40.86% of their 718 respondents believed, incorrectly, that Messenger’s read visibility could be controlled in settings. At the time, it could not.

Nearly everyone knew the feature existed: 92.32% were aware of it. Four in ten were wrong about whether they could switch it off.

The design decision, and the twelve days

WhatsApp turned the second grey tick blue on 6 November 2014. There was no way to turn it off.

By 18 November, twelve days later, an opt-out was being reported, initially only for Android users who installed the app directly from WhatsApp’s site rather than through Google Play. Contemporary coverage described the lack of an off switch as having “drawn the ire of users that wish to retain the privacy of not allowing senders to know when their messages have been read.”

And the opt-out came with a condition that is the real design decision here: turning receipts off means “the user will also not be able to see if the people that the user has sent a message to have read it.”

You can only stop showing if you also stop seeing. Signal states this as a principle rather than a consequence. Its own documentation says plainly: “Both you and your contact must have it enabled to see this status,” and its 2017 launch post adds, “Our implementation of read receipts is simple. There is only one setting.”

That symmetry is a choice, not a technical necessity. A platform could let you see without being seen. None of the major ones does.

Where you can turn it off, and where you cannot

Platform	Read receipts	Can you disable them?
iMessage	Yes	Yes, globally and per conversation
Signal	Yes, opt-in	Yes. Both people must have it on for either to see it
WhatsApp	Yes, blue ticks	Yes, but not in group chats
Messenger	Yes	Only since a test that began November 2023
Instagram	Yes	Only since the same 2023 test
Slack	No sender-facing read indicator	n/a

Two things are worth noticing. For roughly its first decade, Messenger’s “Seen” was mandatory, there was no setting, which is why four in ten users looking for one could not find it. And WhatsApp’s group-chat exception means the largest, most socially exposed conversations are the ones you cannot opt out of.

A note on sourcing: Apple’s and Signal’s own documentation is quoted directly above. WhatsApp’s and Meta’s help centres block automated retrieval, so the WhatsApp group-chat rule comes from the peer-reviewed 2017 paper and the reciprocity rule from 2014 trade press; the Messenger and Instagram opt-out dates come from contemporaneous technology reporting. Check the current setting in the app rather than trusting a table.

What this article will not claim

That being left on read harms you. We looked specifically for experimental, longitudinal or review evidence linking read receipts to anxiety, depression or wellbeing outcomes, and found none. What exists is cross-sectional self-report of momentary reactions: how people said they felt, once, in a survey. That is not a measure of harm, it is not longitudinal, and it is not causal. The 42.8% figure above describes a feeling reported in a questionnaire, and it should not be upgraded into anything stronger.

That this is a form of ostracism. There is a real experimental literature on social exclusion. It has not, as far as we could find, been applied to read receipts, and borrowing its authority would be a category error.

That workplace research settles it. There is a well-validated construct called workplace telepressure, the preoccupation with responding quickly to work messages, with documented links to burnout and sleep quality. It is about employees and email. Extending it to a friend not answering your text is an analogy, not a finding.

That the survey numbers are population estimates. The 718-person study was conducted on Mechanical Turk. A second study reporting that 36% felt pressured to respond once a message was marked read (44% of women, 24% of men) is a self-described student project on 108 humanities students at one Danish university, whose authors say it cannot be generalised further. We cite it as corroboration of direction, not as a rate.

What to actually do

If receipts bother you, understand the trade. On every platform that offers an off switch, switching off costs you the same information you stop giving. That is the deal, and it is worth taking deliberately rather than discovering it afterwards.

Know that “Read” tells you nothing. It reports that a screen displayed a message. It does not report attention, comprehension, intent, or availability, and the evidence says a large fraction of people open messages specifically to send an acknowledgement without replying, while a larger fraction avoid opening precisely so the signal does not fire. Either way, the thing you are reading into it is not there.

And notice which way the feature points. A read receipt gives the sender information about the recipient. There is no reciprocal signal telling the recipient how urgently the sender needs an answer, or whether it can wait until tomorrow. The asymmetry is the design.

Read receipts sit inside a wider set of interface decisions covered in the persuasive design guide. For what a count of other people’s behaviour does to your own judgement, see what visible metrics do to your decisions. For the cost of the interruption itself rather than the signal, what notifications actually do.