Key takeaways<\/h2>\n\n\n\n\n- Replies go out as the brand, not as a person.<\/strong> Three writing styles read to a customer as one erratic voice.<\/li>\n\n\n
- DM threads carry no context of their own – no subject, no history header – so handover context must be written down or it is lost.<\/li>\n\n\n
- There is no draft state and no recall. A send is final, which raises the cost of answering in a hurry.<\/li>\n\n\n
- On most channels the customer can see the message was read, so opening a thread to triage it is not free.<\/li>\n\n\n
- The reply window belongs to the brand, not to the shift.<\/strong> A handover at hour 20 inherits four hours, not twenty-four.<\/li>\n\n\n
- The metric changes from \u201cdid I reply?\u201d to \u201cdid anyone?\u201d, and it is usually not being measured when that happens.<\/li>\n\n\n
- Every fix here is a written agreement, not a feature.<\/li>\n\n<\/ul>\n\n<\/div>\n\n\n\n
Five Things Email Gives You That DMs Do Not<\/h2>\n\n\n\n
Each one is a small absence, and each becomes a process once two people share the inbox.<\/p>\n\n\n\n
No Sender<\/h3>\n\n\n\n
An email from a shared address still carries a name at the bottom. A DM carries nothing – the reply is simply the brand speaking. From the customer’s side there is one continuous voice, so three people writing in three registers does not read as a team, it reads as one person who is warm on Monday and curt on Thursday.<\/p>\n\n\n\n
This is the first thing to fix, and the fix is cheap: a written voice note with three real examples of a good reply. Not a style guide – three examples. People copy examples and ignore guides.<\/p>\n\n\n\n
No Subject, No Thread Context<\/h3>\n\n\n\n
Email arrives with a subject line and usually the whole history quoted underneath. A DM thread is a flat list of messages with no summary, which means picking up somebody else’s conversation requires reading all of it and guessing at what was promised verbally.<\/p>\n\n\n\n
The fix is a convention: whoever handles a conversation leaves a short note in it before stepping away. One line – what was promised, what is outstanding. This is the single highest-value habit on this page.<\/p>\n\n\n\n
No Drafts, No Recall<\/h3>\n\n\n\n
There is no outbox to catch a mistake and no unsend worth relying on. A half-finished reply sent by accident is simply sent, in public-facing tone, under the brand’s name.<\/p>\n\n\n\n
Saved reply templates help more here than they appear to – not because typing is slow, but because a reviewed sentence sent quickly beats an unreviewed one sent quickly.<\/p>\n\n\n\n
They Can See You Read It<\/h3>\n\n\n\n
On most channels the customer knows the message was opened. That makes triage visible: browsing the queue to see what is there starts a clock in the customer’s head that your rota may not be ready for.<\/p>\n\n\n\n
Two practical consequences. Open a thread when you intend to deal with it, not when you are surveying. And if you must open and leave, send a holding line – the read receipt has already made the promise.<\/p>\n\n\n\n
The Window Belongs to the Brand<\/h3>\n\n\n\n
The 24-hour Meta window runs from the customer’s last message, regardless of who was on shift when it arrived. Somebody taking over at hour twenty inherits four hours, not a fresh day, and nothing in the interface shouts about it.<\/p>\n\n\n\n
This is a rota problem rather than a tooling one, and it is the reason handovers need to carry the age of a conversation, not just its content.<\/p>\n\n\n\n![\"Five](\"https:\/\/dmpro.brandid.app\/blog\/wp-content\/uploads\/2026\/10\/team-inbox-social-media-changes.jpg\")
Social DMs are not email, and these five differences are why.<\/figcaption><\/figure>\n\n\n\nThe Metric Quietly Changes<\/h2>\n\n\n\n
And the change is usually noticed a quarter late.<\/p>\n\n\n\n
From \u201cDid I Reply?\u201d to \u201cDid Anyone?\u201d<\/h3>\n\n\n\n
A single person always knows the state of the inbox, because they are the inbox. Two people each know their own half, and the gap between those halves is where conversations go unanswered – each person reasonably assuming the other had it.<\/p>\n\n\n\n
The number that matters is no longer individual throughput, it is unanswered conversations older than your target. Start measuring that on the day the second person joins, not when somebody complains.<\/p>\n\n\n\n
Duplicate Replies Are the Other Half<\/h3>\n\n\n\n
The mirror image of the gap: two people answering the same customer, sometimes differently. It is embarrassing in a way the customer remembers, and it is purely a coordination failure.<\/p>\n\n\n\n
Both problems – the gap and the duplicate – come from the same missing thing, which is a visible answer to \u201cwho has this?\u201d. That is the subject of assigning conversations to a team<\/a>.<\/p>\n\n\n\nMeasure the Same Way Everywhere<\/h3>\n\n\n\n
If reply time is computed differently per channel, the team will optimise the channel with the kindest definition. One definition across the whole queue is part of what the unified social inbox<\/a> is for.<\/p>\n\n\n\n![\"One](\"https:\/\/dmpro.brandid.app\/blog\/wp-content\/uploads\/2026\/10\/team-inbox-social-media-agreements.jpg\")
Habits are private, and a second person cannot inherit one.<\/figcaption><\/figure>\n\n\n\nFive Agreements Worth Writing Down<\/h2>\n\n\n\n
Each takes under an hour and replaces something currently held in one head.<\/p>\n\n\n\n
Five Things Email Gives You That DMs Do Not<\/h2>\n\n\n\n
Each one is a small absence, and each becomes a process once two people share the inbox.<\/p>\n\n\n\n
No Sender<\/h3>\n\n\n\n
An email from a shared address still carries a name at the bottom. A DM carries nothing – the reply is simply the brand speaking. From the customer’s side there is one continuous voice, so three people writing in three registers does not read as a team, it reads as one person who is warm on Monday and curt on Thursday.<\/p>\n\n\n\n
This is the first thing to fix, and the fix is cheap: a written voice note with three real examples of a good reply. Not a style guide – three examples. People copy examples and ignore guides.<\/p>\n\n\n\n
No Subject, No Thread Context<\/h3>\n\n\n\n
Email arrives with a subject line and usually the whole history quoted underneath. A DM thread is a flat list of messages with no summary, which means picking up somebody else’s conversation requires reading all of it and guessing at what was promised verbally.<\/p>\n\n\n\n
The fix is a convention: whoever handles a conversation leaves a short note in it before stepping away. One line – what was promised, what is outstanding. This is the single highest-value habit on this page.<\/p>\n\n\n\n
No Drafts, No Recall<\/h3>\n\n\n\n
There is no outbox to catch a mistake and no unsend worth relying on. A half-finished reply sent by accident is simply sent, in public-facing tone, under the brand’s name.<\/p>\n\n\n\n
Saved reply templates help more here than they appear to – not because typing is slow, but because a reviewed sentence sent quickly beats an unreviewed one sent quickly.<\/p>\n\n\n\n
They Can See You Read It<\/h3>\n\n\n\n
On most channels the customer knows the message was opened. That makes triage visible: browsing the queue to see what is there starts a clock in the customer’s head that your rota may not be ready for.<\/p>\n\n\n\n
Two practical consequences. Open a thread when you intend to deal with it, not when you are surveying. And if you must open and leave, send a holding line – the read receipt has already made the promise.<\/p>\n\n\n\n
The Window Belongs to the Brand<\/h3>\n\n\n\n
The 24-hour Meta window runs from the customer’s last message, regardless of who was on shift when it arrived. Somebody taking over at hour twenty inherits four hours, not a fresh day, and nothing in the interface shouts about it.<\/p>\n\n\n\n
This is a rota problem rather than a tooling one, and it is the reason handovers need to carry the age of a conversation, not just its content.<\/p>\n\n\n\n And the change is usually noticed a quarter late.<\/p>\n\n\n\n A single person always knows the state of the inbox, because they are the inbox. Two people each know their own half, and the gap between those halves is where conversations go unanswered – each person reasonably assuming the other had it.<\/p>\n\n\n\n The number that matters is no longer individual throughput, it is unanswered conversations older than your target. Start measuring that on the day the second person joins, not when somebody complains.<\/p>\n\n\n\n The mirror image of the gap: two people answering the same customer, sometimes differently. It is embarrassing in a way the customer remembers, and it is purely a coordination failure.<\/p>\n\n\n\n Both problems – the gap and the duplicate – come from the same missing thing, which is a visible answer to \u201cwho has this?\u201d. That is the subject of assigning conversations to a team<\/a>.<\/p>\n\n\n\n If reply time is computed differently per channel, the team will optimise the channel with the kindest definition. One definition across the whole queue is part of what the unified social inbox<\/a> is for.<\/p>\n\n\n\n Each takes under an hour and replaces something currently held in one head.<\/p>\n\n\n\nThe Metric Quietly Changes<\/h2>\n\n\n\n
From \u201cDid I Reply?\u201d to \u201cDid Anyone?\u201d<\/h3>\n\n\n\n
Duplicate Replies Are the Other Half<\/h3>\n\n\n\n
Measure the Same Way Everywhere<\/h3>\n\n\n\n
Five Agreements Worth Writing Down<\/h2>\n\n\n\n