We talk a great deal about the roles AI may create. It is easy to imagine a future organisation chart full of unfamiliar titles, and much harder to recognise the change while it is happening. Sometimes the first sign is a familiar job whose centre of gravity has moved.

Social media management may be one of those jobs. In an Adweek interview created in partnership with Hootsuite, Hootsuite CMO Christina O’Reilly described social managers as “architects or directors of AI agents”. It is an arresting phrase. I am more interested in the shift underneath it: if AI helps produce content and watch conversations at scale, what does the person closest to those conversations become responsible for?

Focus On: A familiar job, a wider view

Social managers have never simply pressed Publish. They have answered awkward questions, spotted complaints before they reached a formal report, and learned to tell the difference between a passing joke and a story that is about to travel. AI has not invented their proximity to customers. It may change what they can do with it.

Imagine an assistant that groups thousands of comments and flags an apparent rise in complaints about a product. It has done useful work. Yet the pattern might reflect a fault, a stock shortage, a confusing promotion, or a handful of unusually vocal accounts. The number of mentions alone cannot settle the matter. Someone has to read enough of the underlying conversation to understand it, then look beyond the social feed.

O’Reilly offered a useful illustration. She said that, for an unnamed beauty retailer, a general-purpose AI tool brought up years-old information when asked about current complaints. A system connected to live social data instead pointed to an out-of-stock product. It is an anonymous example from a vendor conversation, but the mistake is recognisable. A fluent answer to the wrong moment is still the wrong answer.

The expanding role begins there. The social manager can ask whether support tickets tell the same story, whether the stock record confirms it, and whether customers are unhappy with the product or with their inability to buy it. A reply might satisfy one customer. Understanding the pattern could help the business respond to many.

On another day, the same person may review an AI-drafted reply that misses the tone of a complaint, or reject a proposed post that follows a trend the brand has misunderstood. The work still contains writing and publishing. More of it may involve deciding when a system has read the room well enough, and when a person needs to take over.

What the work looks like

Samsonite’s experience across 13 markets in Asia-Pacific and the Middle East shows a fuller version of that journey. Its social listening partner, Onclusive, reports that the regional team faced roughly 50,000 mentions a month, across several languages and platforms. Most were promotional noise. The challenge was to find the smaller number that deserved attention elsewhere in the company.

The arrangement paired AI-powered monitoring with people who checked alerts each morning. They did more than forward a digest. Customer service received daily alerts, regional leaders got weekly summaries, and marketing and product teams saw monthly trends. Each group received information at a pace that suited the decisions it could make.

One recurring issue concerned repair delays. Onclusive says the complaints contributed to a new replacement policy with time-bound guarantees. The supplier reports that outcome, so we cannot independently measure how much the listening programme caused it. We can see the chain of work: a public complaint is recognised as part of a pattern, checked by a person, brought to the right colleagues, and reflected in a service decision. The story ends somewhere other than the comments section.

This is the part that makes the role feel different to me. The person responsible for social can still run a channel, but may also help the organisation interpret what customers are experiencing. One task is to decide what to publish next. Another is to decide which conversations reveal a problem worth investigating. The second has consequences for people who may never open the social dashboard.

None of this means social teams have only just discovered customer insight. They have been trying to get the rest of the business to listen for years. AI can make more conversations visible and shorten the time between a signal appearing and someone examining it. That increased reach may give an existing role a broader remit. Samsonite’s case shows the work, although it does not show a new job being created.

A role before a title

Of course, the expansion is not guaranteed. AI could make the content calendar busier while leaving the same person to handle every reply and report. Hootsuite’s commissioned survey of 500 social media managers in the US and UK found that 43% spent more than 11 hours a week using AI tools, while 48% still spent more than 11 hours manually scanning platforms for trends. Time spent with a tool does not tell us whether the work improved, but those figures hardly describe an empty diary awaiting a grand new assignment.

There is also a reason to be careful with the signals themselves. Social conversation is immediate, but it is not a representative sample of customers. The people most inclined to post may be delighted, furious, or simply very online. That is why the human role involves checking what the system has surfaced against other evidence, and knowing when a handful of posts deserves attention even before a pattern can be proved. AI can accelerate discovery; it cannot turn every mention into a reliable conclusion.

Perhaps some organisations will eventually give this work to a dedicated social intelligence specialist. Elsewhere it may be the familiar social manager, now spending less time assembling the evidence and more time working out what it means. Either way, the interesting change is in the responsibility, not the label.

When we talk about new roles created by AI, we often expect them to arrive as vacancies with unfamiliar names. Social media management suggests another route. An existing professional starts carrying a question further than their channel: what are customers telling us while there is still time to respond? The title can catch up later.

Follow me

That’s all for this week. To keep up with the latest in generative AI and its relevance to your digital transformation programs, follow me on LinkedIn or subscribe to this newsletter.

Disclaimer: The views and opinions expressed in Chronicles of Change and on my social media accounts are my own and do not necessarily reflect the official policy or position of S&P Global.