Who Actually Owns the Social Inbox.

Ask a marketing lead who owns the social inbox and you will usually get a tool name back rather than a person’s name.

That answer is the whole problem.

Direct messages, mentions, story replies and tags are a service queue that happens to arrive through a marketing channel. A queue with no named owner drains at whatever speed the busiest person can spare, which is why brands with a perfectly good unified inbox still miss things.

Ownership is the fix, and it costs nothing to decide. One named owner per queue, a cover window you can genuinely staff, and one written escalation route. Do that before you buy anything.

The social inbox is a service queue wearing a marketing badge

Three different jobs land in the same place. Someone asking where their order is needs system access. Someone asking whether you ship to their country needs commercial knowledge. Someone complaining in public needs judgement about tone and risk.

A social manager is usually hired for the third job and sometimes the second. They are rarely equipped for the first, because answering it means opening a system marketing does not have a login for.

So the message waits. Not through negligence, through structure.

Why a unified inbox does not fix it

A shared inbox tells you a message exists. It does not tell any specific person that answering it is their job today.

That distinction is the entire gap. Visibility is a tooling problem and software solves it well. Accountability is an operating problem, and no software has ever solved it, because a tool cannot assign a human being.

The same pattern shows up in comments, where teams often discover that nobody has decided who moderates the comment section either. Inbox and comments fail for the same reason, in the same way, at the same moment.

The three ownership models, and what each one costs

Published guidance for larger teams tends to settle on three shapes. Each buys something at a price.

• Marketing led. The social team answers everything. Brand voice stays consistent and it copes at low volume, but technical questions stall.
• Support led. The support desk owns replies, usually through a connected help desk. Answers are accurate and get tracked, but they read like tickets.
• Shared. Marketing holds engagement and voice, support handles resolution, with handoff rules between them.

The shared model is the one most enterprise guidance recommends, and it is also the hardest to run, because it is the only one with a handoff inside it.

That matters more than it sounds. The first two models fail loudly. The shared model fails quietly, in the gap.

The hour nobody owns

Marketing sees the message first. Support has the answer. Between those two facts sits an unowned hour, and nothing in the tool moves the message across it.

You can spot the gap without a report:

• Messages get read, reacted to, then answered a day later by someone else.
• The same question gets two different answers in the same week.
• Replies stop entirely at 6pm on Friday and restart on Monday morning.
• People forward screenshots of DMs into chat threads to ask what to say.
• Nobody can tell you how many messages arrived last month.

The last one is the clearest signal. An unowned queue is an unmeasured queue, because measurement follows accountability rather than the other way round.

Split the inbox before you assign it

You cannot name one owner for a queue that is actually five queues. Split by the skill the reply needs, not by the platform it arrived on.

• Service and account problems.
• Sales and pre-purchase questions.
• Complaints made in public, on the post or in the mentions.
• Press, partnership and creator approaches.
• Abuse, spam and impersonation.

Most brands find the split is already there informally. Writing it down turns a habit into a rule.

Step 1: Name one owner per queue

An owner is a person, not a department. “Customer service owns service DMs” is not ownership, because a department cannot be asked why something sat for two days.

Name the person, name the deputy, and write both down where the team can see them.

One person may own three queues in a small team. That is fine. What is not fine is a queue owned by everyone, which is the same arrangement as a queue owned by nobody, but harder to notice.

Step 2: State a cover window you can actually staff

Pick hours you can honestly hold, then publish them. A stated window of 9am to 5pm on weekdays, actually met, beats an unstated promise of always-on that quietly becomes twelve hours on a Saturday.

Then decide what happens outside it. Usually the honest answer is an auto-reply that states the window and points to a channel that is staffed, which is a better experience than silence.

Weekend cover is a staffing decision with a real cost, so make it deliberately. Getting the workflow and cover model right is the difference between a rota and a rolling favour asked of whoever happens to be online.

Step 3: Write one escalation route

One route, on one page. Who receives it, how they receive it, and how fast they are expected to move.

Three tiers is enough for most brands:

1. Standard. The queue owner answers within the stated window.
2. Escalated. Anything needing account access, a refund decision or a policy answer. Named receiver, named channel, a stated target.
3. Urgent. Legal risk, safety, press, or a complaint gaining traction in public. Goes to a named person immediately, by phone if needed, not by leaving a note in the tool.

The point of writing it is not the document. It is that the person holding the message at 4:55pm on Friday knows exactly where to put it.

The published benchmarks disagree, so set your own

Response time targets get quoted with a confidence the research does not support. One widely cited annual consumer survey puts it at roughly three quarters of consumers expecting a reply within 24 hours or sooner. Other published guidance for large teams treats one hour as the working standard. Some roundups restate that same survey at a slightly different figure than the survey’s own write-up carries.

The numbers also move by platform and by message type. A complaint is not a stockist question, and people wait differently for each.

What is consistent across the research is direction rather than duration. Fast replies to direct questions rate among the most appealing things a brand does on social, and a large share of social users say they will simply buy from a competitor when a brand does not respond at all.

So set a target you can meet, measure against it, and tighten it. A met four hour target is a stronger operating position than a missed one hour target.

What good looks like

A message arrives. The tool routes it to a queue. The queue has a named owner who is on shift, because the window is staffed rather than assumed.

The owner answers or escalates inside a stated time, and escalation is a route rather than a favour. Anything with reputational weight moves straight to the people who handle a social media crisis, without anyone pausing to work out who to ask.

At the end of the month someone can say how many messages arrived, how many were answered inside the window, and which queue was slowest. That last sentence is the one most brands cannot say today.

How NBK thinks about the social inbox

Social operations is the lane NBK cares most about, and the inbox is where a missing system shows up fastest. Content problems take a quarter to reveal themselves. An unowned inbox reveals itself the first time someone important is ignored.

The order matters too. Decide the owner, the window and the route first, then choose a tool to carry them. A tool bought before those three decisions just gives the confusion a nicer interface.

None of this requires headcount for most brands. It requires someone senior to make three decisions and write them down.

Next step

If your team is answering messages but nobody can say who is responsible for the ones that go unanswered, the constraint is ownership rather than effort. NBK can help build the system behind it, starting with who owns what, and when.