What running a content team looked like 10 years ago - and what it looks like now
About 12 years ago, I ran a content operation that looked like most content operations at the time. A room of writers. An SEO lead feeding them keyword briefs. Templates, workflows, a publishing policy, someone to handle distribution. If you wanted ten blogs a week, you hired ten writers. If you wanted to post on social, you hired someone to do that too. Some of those writers were composing in HTML because the CMS didn't have a rich text editor. It was manual all the way down.
It worked. Slowly, expensively, and entirely dependent on headcount.
I've spent the last few years running something that produces comparable output with a fraction of the people. This is the actual operating model, at Backbase, where we've tripled our commercial and informational traffic and are doing better in our category than we ever have. The system now handles work that previously required a room of humans.
What changed
The production logic flipped. In the old model, the human was the engine - researching, writing, linking, formatting, distributing. The system existed to coordinate the humans. In the current model, the system is the engine. The human exists to feed it the right inputs and catch what it gets wrong.
What those inputs need to be is specific and non-negotiable. You need a strong knowledge base and proprietary data - real transcripts, from real conversations - so the model is writing from your thinking, not blending the internet back at you. Strong proprietary source material produces something that sounds like it came from someone who genuinely knows the subject.
Internal linking, external linking from a vetted source list, formatting - all of that runs automatically. The linking doesn't hallucinate if you force the agent to scrape and read the source it's citing before it includes it. That is correct workflow design.
The result: ten to twenty pieces of content a week, one good editor, and a human in the loop who makes sure nothing slips through. That's the whole operation.
What this means for the people doing this work
The content manager who thrives in this environment is a producer who understands systems and editorial judgment. They understand what good content looks like, they know how to brief a system, they can build and maintain a knowledge base, and they have the editorial judgment to know when the output needs work. Systems thinking is the primary skill required.
The role requires systems thinking and editorial judgment - including the ability to direct an agentic workflow and understand what it can do. You don't need to build the agentic infrastructure from scratch to benefit from it. You need to understand what it can do and how to direct it.
There's a lot of noise right now about AI content dying - that Google will penalise it and that audiences will reject it. Some of that is true of the bad version. Strong inputs and rigorous human review produce content that performs.
What I've seen work is the version where the system handles volume and the human handles quality control and strategic direction. That combination works, with one person who genuinely understands both the subject matter and the workflow.
The operation has changed, and the quality standard remains constant. That is the point.