Contact names against phone numbers: what a message export should show
A message export that says only “Mom” at the top names nobody, and one that says only
+1 555 0142 names somebody nobody recognises. A document meant for a court or a lawyer wants
both: the name, so a reader can follow the conversation, and the handle it belongs to, so the
name can be checked against something. The PDF carries both, the name on every message and the
handle on its front matter, and the one thing no format will do is let a label stand where the
handle should be.
This is the part of the method that how to export iPhone messages for court describes from the other end, as one of the things a reader of your document will ask about.
Where does the name in an export come from?
From the address book on the phone, which travels inside the backup. A conversation with somebody in your contacts exports under the name your contacts hold. A conversation with a handle that was never saved exports under the handle, because there is nothing else true to print.
One wrinkle that catches people out: an iMessage thread is not always addressed to a phone number. It can run to an Apple Account email address instead, and then the handle printed beside the name is that address. It is still the checkable part, and still not something to replace with a label.
That is why two exports of the same person’s messages, made from two different phones, can show two different names. Neither is wrong. Both are reporting what one phone’s address book said, which is a fact about that phone rather than about the person.
A group conversation works differently: the group’s own name is used when somebody typed one, every member is listed on the front matter, and no contact photograph is shown, because picking one member’s picture would be an arbitrary choice.
Why is a contact name a problem on a document?
Because an address book is a private nickname store and a document is not. The entries in it were written for your own convenience: “Mom”, “Work”, “Do not answer”, a first name with no surname, an initial, a nickname only you understand.
Printed on a filed document, those labels do three unhelpful things at once. They tell the reader nothing checkable. They can look like an editorial opinion about the other party. And on a thread with somebody who is not in your contacts, they disappear entirely and leave a bare number where the reader expected a person.
The design answer is the one worth copying whatever tool you use: the name is a label, and the handle is the part a reader can check. Keep them both, keep them visibly separate, and never let a label stand where the handle should be.
How do you put the right name on it?
You correct the label before exporting, on the export screen, and the document then records that you did. The correction replaces the printed name and changes nothing else: the number stays exactly as the backup holds it.
What the document does with a correction is the part that makes it usable rather than suspect.
On a document where at least one name was corrected, the integrity page gains a block headed
Participant names, listing each corrected name with its handle beside it and what the
address book had held before, as either address book: Mum or no address-book entry. Under
the list sits this sentence:
The names below were confirmed by the person who produced this document and replace the ones held in the iPhone’s address book. The number beside each name comes from the backup and was not changed.
That is the honest shape of a renaming feature: the edit is disclosed, attributed to you, and printed next to the thing it did not change. On a document where nothing was corrected the block does not appear at all, rather than appearing empty, because a heading suggesting that somebody could have been renamed unseen is worse than no heading.
What a correction will not do
Four limits, and the first two are refusals the program makes on purpose, because the alternative would put something untrue on the page:
- It will not be applied to a handle this conversation does not hold. A name typed for somebody who is not in this thread changes nothing, and the integrity page does not list it. A disclosure of an edit that never landed is still a false statement.
- It will not rename a group. A group’s name was typed by a person on a phone, so correcting one member’s label leaves the group’s own title alone. Nobody would guess that fixing a name had silently retitled their document.
- It will not stick if you clear the field. A blank name means “leave the address book alone”, which is how you undo a correction.
- It will not change the number anywhere. Not on the cover, not on the integrity page, not in the data file.
Which formats show the name and which show the number?
The readable formats show the label you confirmed; the CSV keeps the raw handle. That split is useful rather than an oversight, and it is worth knowing before somebody opens the spreadsheet and asks why the names are missing.
| Format | Sender appears as |
|---|---|
| The display name, with the handle on the front matter | |
| TXT | The display name |
| HTML | The display name, the PDF’s document saved as a file |
| CSV | The handle itself, or me for your own messages |
| RSMF | A container for a review platform rather than a document to read |
A data file full of labels would be a data file you cannot join to anything, so the handle is
the right value for the sender column, and the readable document is where the reader meets
the name. If you need both in one place, export the PDF and the CSV from the same run: they
describe the same conversation and the front matter ties the labels to the handles.
What none of this proves
The document can say which number sent which message, and that you confirmed whose number you believe it to be. It cannot establish that the number belongs to that person. Nothing read out of a phone backup can: the link between a number and a human being lives with a carrier, in a contract, or in that person’s own admission.
This matters because it is a gap an opposing side can reach for, and the answer is not a better export. Whose handle it is gets established by whatever route the proceeding allows, and that is a question for whoever is advising you rather than for this page: the content of the messages, other correspondence with the same handle, what the other party itself accepts, and records held elsewhere are all things that get used, in combinations that differ by country and by case. A document that implied it had settled the question for you would be doing you harm, which is why the integrity page says “confirmed by the person who produced this document” rather than anything stronger.
What courts tend to ask for around this is in text messages as evidence, what courts require.
Questions people ask
Should I put the real name or the phone number on an export for court?
Both, and the PDF is the format that prints both without being asked: the name on every message, the handle on the front matter. If the address book holds a nickname, correct it to the person’s actual name as the proceeding uses it, and let the document print the handle and the old label beside your correction. A document headed “Mom” raises a question you would rather not spend time on.
Can I hide the other person’s phone number?
Not through the naming feature, which exists to keep the handle visible. If a handle genuinely must not appear, that is a redaction decision to take with whoever is advising you, and a redaction is something you do deliberately and disclose, not something a tool should do quietly on your behalf.
Two consequences follow, and they are the practical half. A redaction is done to the exported
file, so the checksum that travelled with the original no longer describes the version you
hand over: keep the original export and its .sha256 file, and say which one is which. And a
group conversation carries the handles of people who are not party to anything, which is worth
a thought before the whole thread goes out.
Why does one conversation show a name and another show a number?
Because one of those people is in the phone’s address book and the other is not. It is not a failure of the export, and correcting the second one before exporting is exactly what the naming feature is for.
Does any of this apply to a WhatsApp export?
Partly, and the difference is worth knowing before you rely on it. A WhatsApp export file carries the names the exporting phone wrote into it, not handles from an address book, so there is often no number to print beside a corrected name. The label can still be corrected; what cannot happen is the part this page calls the checkable half, because the file never held it. That is one of the reasons a conversation read out of an Apple backup and a conversation read out of an export file are not the same kind of record.
Does changing a contact on my phone change an old export?
No. An export is a file written at one moment from one backup. Changing a contact now changes what a future export of a future backup would print, and nothing about the document already on your disk, whose checksum would no longer match if it did.
Where to go next
- How to export iPhone messages for court, the hub this page hangs off.
- Text messages as evidence, what courts require, for what gets asked about a filed conversation.
- Exporting iPhone messages with their metadata, for the rest of what travels beside the words.