Issue 22

If you’ve bumped into LinkedIn discussions about content engineering and artificial intelligence (AI) lately, they might leave you confused or frustrated, so let’s take a step back to see where the term came from, what it’s being made to mean now, and what’s at stake in that shift.

The roots: content engineering before AI

Ann Rockley first discussed what became content engineering in Managing Enterprise Content (2002), introducing the idea of intelligent content, which structurally rich and semantically categorized, and is therefore automatically discoverable, reusable, reconfigurable, and adaptable.

In 2010, Joe Gollner made the case for something called content engineering as separate from and complementary to content strategy. (The post on his site is no longer available.)

The term rose in prominence in 2013, when Rockley gave the closing keynote of the LavaCon conference and Mark Baker blogged about it on Every Page Is Page One. As Baker put it:

​Cruce Saunders refined the definition in 2016, describing content engineering as organizing the shape and structure of content by deploying content models in authoring and publishing processes to meet organizational requirements and using technology such as CMS (content management systems), XML (extensible markup language), Schema, AI, APIs (application programming interfaces) and others.

He identified six core disciplines:

• Model (content modeling)
• Metadata
• Markup
• Schema
• Taxonomy
• Graphs

At its core, content engineering is focused on content structure, semantics, and the systems that enable that.

Val Swisher and Regina Lynn Preciado published The Personalization Paradox (2021), which highlights how content engineering enabled dynamic content outputs long before AI hype hit us. “Dynamic outputs are relatively new to the publishing scene and require a lot of coordination, collaboration, and content engineering,” they wrote.

​In June 2021, I published an article that resonated widely. I contextualized content engineering alongside content strategy, content design, and content operations. (I revisited that post in Issue 13 on its anniversary and have done deeper dives in Issue 20 and Issue 21.)

In January 2023, I published another article, predicting the rise of content engineering.

“There are not enough content engineers in the world,” Saunders said in an interview for that article. “The demand for content engineering continues to grow, and there is a shortage of qualified individuals to fill this need.

​Paola Rocuzzo chimed in after my 2023 article, reinforced this and pointing out that folks working deep in XML and other markup standards used the title content engineer in the last 30 years.

​Colleen Jones summarized the discipline in Content Science Review (July 2025): “Content engineering is a discipline that focuses on designing, structuring, delivering, and managing content in a way that is strategic, scalable, and workflow-optimized.”

She went on, saying it “isn’t about writing, creating, or even managing content. It’s about applying systems thinking, engineering rigour, governance, and technology to manage content as a strategic business asset.”

Jones highlighted several benefits of content engineering. It helps teams:

• Scale content consistently
• Maintain content quality
• Enable content automation and AI
• Drive content ROI and efficiency

Content engineering employs a variety of ways of working to do this, Jones wrote:

• Content modeling and metadata strategies
• Templating and component-based design
• Automation and tooling development
• Governance and operations maturity
• Analytics and content intelligence (insights into content effectiveness)

Even before AI owned the hype cycle, content engineering was foundational for scaling, personalization, and sophisticated content systems. Without it, dynamic outputs and enterprise-level content management would be far more onerous, and users would not have content targeted to their needs or interests.

The remix: marketing’s new definition

The recent discussion of content engineering on LinkedIn is driven heavily by marketers.

Josh Spilker of the AI search tool AirOps published an article The Rise of the 10x Content Engineer in February 2025. His “10x content engineer” has the following characteristics:

• Thinks in systems
• Blends AI and automation to grow content production with an eye on quality
• Creates repeatable workflows for programmatic SEO, content refresh cycles, and personalized user experiences
• Moves beyond traditional content formats, leverages data-driven insights, AI-generated assets, and dynamic content hubs

As such, Spilker said a content engineer is a combination of product manager, content strategist, writer/creator, AI prompt engineer, and SEO manager.

However, Spilker’s colleague Alex Halliday described a content engineer as focusing on workflow orchestration (part of what is known otherwise as content operations), AI model optimization, and quality assurance.

Loreal Lynch, a chief marketing officer, wrote an article in September 2025 titled The Rise of the Content Engineer: Redefining Marketing in the Next Era. In it, Lynch explained content engineering as “designing, building, and optimizing content production systems powered by AI.”

Content engineers, Lynch said, “design, orchestrate, and govern AI-powered content systems that can scale quality, consistency, and personalization across an enterprise.”

Lynch continued: “While the Content Engineer may be an evolution of traditional content roles, there is a key distinction: A content marketer produces pieces. A content engineer produces systems that produce pieces at scale—AI-powered, data-driven, and adaptable.”

Content engineering is transforming marketing in four ways, according to Lynch:

• Systematizing brand voice consistency
• Orchestrating cross-channel content ecosystems
• Architecting feedback loops
• Scaling personalization without losing strategy

Content marketing director Ryan Law spiced things up with his September 2025 AHREFS blog post I Wouldn’t Hire a Content Engineer, and You Shouldn’t Either. Law described his understanding of content engineering this way:

While there’s clear overlap between the definitions, the differences in focus reveal deeper tensions between marketing and non-marketing content disciplines.

Where the marketing-driven view of content engineering overlaps with the traditional definition

Both views have a strong overlap in their understanding of content engineering. In both perspectives, some themes surface:

• Content engineers must be systems thinkers.
• A focus for content engineering is scaling content.
• AI is part of the picture.
• Content engineering enables improved workflows.
• Content engineering supports content strategy efforts.
• Omnichannel publishing requires content engineering. I have a hunch that traditional content engineering understands this in a totally different paradigm than marketing-driven content engineering does.

Where the marketing-driven view of content engineering diverges from the traditional definition

While the two views of content engineering seem on paper to be similar, it’s pretty easy to come away thinking that the new view is drastically different in application.

For instance, ContentEngineer.io has a footnote that seems to view traditional content engineering as something in the past.

It seems that you can divide the differences in application of the definitions in several ways.

• Semantics v. automation: Traditional content engineering focuses on meaning; marketing content engineering focuses on mechanisms.
• Architecture v. production: Traditional content engineering designs frameworks; marketing content engineering builds workflows.
• AI as a tool v. AI as prerequisite: For traditional content engineering, AI is a tool or a piece of the larger system; for marketing content engineering, AI is the centerpiece.
• Broad scope v. narrow application: Traditional content engineering focuses on enabling content across an enterprise and connecting isolated parts of an organization. Marketing content engineering seems focused on campaigns.

Both definitions are valid within their own contexts, but when we use the same term for such different skill sets, we risk confusing teams, mis-hiring, and undervaluing the semantic foundations that make AI-driven content possible in the first place. Goodness knows, we already face that in content strategy!

Why this semantic drift matters

If you accept that the application of the definitions differs, you can start to see why that’s problematic.

• Misaligned hiring expectations: It seems quite likely that someone would hire a prompt engineer when they really need an information architect.
• Fragile systems: It’s not hard to see how AI tools built with narrow focus will lack broader integrations or the accuracy and efficiency enabled by structured, semantic content foundations.
• Loss of meaning: When terms get redefined in related-but-separate disciplines, they lose their meaning (as already mentioned). Confusion reigns as groups inevitably need to collaborate, yet they talk past each other.

However, I think there’s also an opportunity here. The attention from marketing could draw new people into content systems thinking, but I think we need to keep the conversation constructive.

Scaling, personalization, and AI: where traditional content engineering still shines

Law discredited the importance of scaling content by focusing on the SEO value of mass amounts of published content.

In a sense, there’s wisdom in this. Content strategists have long known that having too much content is hard to maintain and can dilute SEO.

But scaling content is much more than this. It’s about large enterprises with legitimate needs for thousands and tens of thousands of pieces of content that are compiled in different ways for different channels and different audiences.

In the simplest terms, how does a company get to a content ecosystem with tens of thousands of webpages? Someone has architected the structure of content to allow this. And you get more value from that content structure if a traditional content architect has been involved, applying traditional content engineering practices.

But start drilling down. Scaling content means that for those tens of thousands of pages, metadata and taxonomy may be driving the internal linking, removing the need for marketers and technical writers and producers (and so on) from manually maintaining internal links. That’s content engineering.

Drill down again. On those tens of thousands of pages, do any have personalization—at any granularity? While you can do some personalization manually, eventually you hit a scale problem. Personalization makes the content ecosystem exponentially more complex and creates an explosion of content variants to manage.

There’s a sense that AI can shortcut that manual personalization by brute force, but the reality is that well-engineered content can automate the personalization without the intensive computing power of AI. (That said, I think creating the content variants required by personalization is a good AI task.)

AI may feel like the next great leap forward, but content engineering is what makes that leap possible. It underpins everything AI does with content, improving efficiency, accuracy, and relevance. After all, the content that AI draws from has to come from somewhere, and without solid models, metadata, and structure, those AI tools are just guessing.

In that sense, content engineering isn’t being replaced by AI. It’s giving AI something meaningful to work with.

The future of the team

​Rahel Anne Bailie wrote yesterday that “skills will shift.” But we can’t skip the foundations—metadata, modeling, taxonomy, governance. Otherwise, we’ll just be teaching AI to replicate our current chaos at scale.

So maybe the way forward isn’t choosing between traditional and marketing versions of content engineering. Maybe it’s about building content teams that draw from all parts of an organization where both perspectives coexist and collaborate—where content engineers, content designers, prompt engineers, marketers, producers, and technologists each play to their strengths and truly 10x their organization’s content experiences.

These things can all be true

Building the team of the future is exciting, but it’s only part of the story. To understand the full picture, we need to recognize that multiple truths about content, AI, and roles exist simultaneously.

Here’s my effort to start unpacking the nuance of these multiple truths:

AI capabilities v. human oversight

• AI can generate personalized content at scale.
• We still need humans to exercise judgment for brand voice, accuracy, and meaning.
• AI doesn’t replace the need for content governance. It amplifies gaps
• AI has lots of promise for helping improve content quality.

Scaling up v. scaling down

• Marketers and content teams need to scale up their content.
• Businesses need to manage unprecedented amounts of content.
• Marketers and content teams need to curate and cull their content.
• Businesses need to know what content they have and know its value.

Structured v. unstructured

• Metadata, taxonomy, content models are more important now than ever.
• Marketers and content teams need one-off content. It’s unavoidable.
• Marketers and content teams need to be able to experiment with content.
• In some cases, experimentation needs to be bespoke.
• In some cases, experimentation can be aided by structure.

Content roles

• Marketers are the experts in content for marketing goals.
• Content strategists and content designers are the experts for understanding and shaping organization-wide content experiences.
• Existing roles will change with the adoption of AI.
• There are gaps in content-related roles.
• New positions and job titles are needed for these gaps.

Some of this seems contradictory—but that’s reality. The question is, how do we build teams and systems that can handle it?

John Collins

Thanks for reading! Did someone forward you this email? Subscribe here​ If you’re already a subscriber and you found value in something here, tell your friends and colleagues to subscribe!

Welcome to the 6 new subscribers who joined us since the last issue of Model Thinking.