paidAffiliate disclosure. The partner link in the masthead and in the band beside the copy on this page is a sponsored link to a partner operator, and this site may be paid if you open an account through it, at no extra cost to you. It carries rel="sponsored noopener" and opens in a new tab. A desk about the messages an operator sends has no business leaving its own material unsaid, so its interest is stated here rather than left for a footer: one link funds the site, and no operator, platform, mail provider or product is named, rated or recommended anywhere on it.
The Mailroom / The proof
Fifty sent, forty-eight arrived, twenty-one read
The proof: what shows that a message arrived
Three kinds of evidence are routinely described by one word. On the samples the operator’s own log covers all fifty messages, a delivery receipt covers forty-eight, and a read receipt covers twenty-one - and the reader can only ever obtain the third.
Desk spec
- sent
- 50
- arrival evidence
- 48
- read evidence
- 21
- no evidence at all
- 2
the five kindsA receipt records something that already happened, marketing promotes something, a service note explains a state of the account, a security alert says something changed, and a legal notice is the message whose arrival is itself a step.
the two switchesOne stops marketing and nothing else - 12 of 50 messages on the samples. The other does not stop anything: it changes the channel the remaining 38 travel on, which changes how they arrive and what they prove.
the three countsSent is what the operator's log holds (50 of 50), arrived is what a delivery receipt shows (48 of 50), and read is what a read receipt shows (21 of 50). Three counts, routinely called one.
Direct answerThree kinds of evidence exist. The operator's send log covers every message it sent; a delivery receipt covers 48 of 50 on the samples; a read receipt covers 21 of 50. The two posted messages have no arrival evidence at all, and the only evidence a reader can produce without asking the operator is the message itself.
Three counts that get called the same thing
The gap between them is where most disputes about "I was told" actually sit, and it is arithmetic rather than argument.
Sample F - three kinds of evidence across the fifty messages
| Evidence | Covers | Share | What it shows | Who holds it |
| the send log | 50 | 100.0% | that a send was attempted | the operator |
| a delivery receipt | 48 | 96.0% | that the channel carried it | the operator |
| a read receipt | 21 | 42.0% | that someone opened it | the operator |
| three kinds | 50 / 48 / 21 | 100% / 96.0% / 42.0% | one of the three is not evidence of arrival | none of them is held by the reader |
sample F - the three counts, and the gap between them
sent (log) = 50 -> 50 / 50 = 100.0%
arrival (receipt) = 48 -> 48 / 50 = 96.0%
read (receipt) = 21 -> 21 / 50 = 42.0%
messages with no arrival evidence = 2, both by post (2 / 50 = 4.0%)
of the 48 that arrived, read evidence covers 21 -> 21 / 48 = 43.8%
so the operator can show a send for twice as many messages as it
can show a reading, and the difference is not a defect - a reader
is entitled not to open a marketing message.
What the reader can hold
The first two kinds belong to the operator. The third belongs to whoever opened the message, and the reader's own copy - the message as received - is the only artefact they can produce unaided.
sample E and F - the chain the reader can build
the message as received : the reader's own copy
the receipt of the event it describes : a second, independent copy
the date on the message : the only timestamp the reader sees
3 items, all of them obtainable without asking the operator
against 3 kinds of evidence, 2 of which only the operator holds
so a reader who keeps the message and its receipt has half the
chain, and the half the operator holds audits the other half.
How to use the three counts
When a message is disputed, the useful question is not "was it sent" but "which of the three counts is being relied on" - because only one of them is about arrival at all.
- Ask which of the three is being relied on: the log, a delivery receipt, or a read receipt.
- Treat a send log as evidence of an attempt, and say so if it is offered as proof of arrival.
- Keep the message and its receipt together, because the pair is the reader's own copy of the chain.
- Expect no arrival evidence on a paper channel, and prefer a channel that leaves one for anything time-critical.
- Do not expect a read receipt to exist for a marketing message - the absence is not evidence of a problem.
Read next