Contently: The operating model behind trustworthy content at scale
Scaling content production is a problem most teams approach through tooling: add an AI writing assistant, increase output targets, see what holds. Alex Soto’s May 2026 piece on Contently’s Content Strategist blog argues that the tooling question is secondary to an organizational question—and that teams skipping the organizational layer produce more content, not better content.
The article introduces a four-layer operating model for content programs that need to scale without losing credibility, particularly relevant for financial services, healthcare, and other regulated industries where factual errors carry compliance consequences.
Layer one: vetted creator networks
The model begins with the people producing content, and the argument here is precise: anonymous or AI-only content fails both search engines and compliance teams. Google’s 2025 updates penalized what it called “scaled content abuse”—high volumes of content without verifiable expertise behind it. AI Overview systems, meanwhile, preferentially cite named, credentialed sources. The practical implication is that content teams need human subject-matter experts who are identifiable and auditable, not interchangeable contractor pools producing generalist drafts.
Layer two: structured workflow
Soto describes a five-stage process: brief, source, draft, review, publish. The value is not the stages themselves—most content teams have some version of these—but the mandatory checkpoints at each transition. Without checkpoints, Soto observes, editors end up doing project management rather than editing: chasing sources, correcting scope drift, and fixing voice inconsistencies that would have been caught earlier. The audit trail that a structured workflow creates also matters for regulated industries where documentation of review steps is a compliance requirement.
Layer three: AI within guardrails
AI enters the model at specific stages rather than as a general replacement for human judgment. Research synthesis, draft scaffolding, and metadata generation are the examples given—tasks where AI accelerates work without substituting for the expertise that makes the content credible. The article is clear about what remains off-limits: final byline voice, factual claims in regulated content, and any output that goes to publication without human review. The framing is AI as an accelerant for particular steps, not as a parallel content production system.
Layer four: governance and measurement
The fourth layer addresses what the team measures and how it adjusts. Soto argues for tracking voice consistency, editorial pass rates, and share-of-voice in AI Overviews alongside or instead of raw traffic metrics. These measures create feedback loops that inform the other three layers—identifying which creators produce content that needs heavy editing, which workflow stages generate the most rework, and which AI prompts drift from brand voice.
Who this is for
The article is aimed at VPs of Marketing, content directors, and compliance leaders managing programs at meaningful scale. Smaller teams will find the framework useful as a diagnostic tool even if the full operating model is more than their context requires. The piece works well alongside Soto’s earlier writing on AI search citations, since both address the same underlying question: how do content programs stay relevant when AI systems are increasingly the first place audiences receive answers.