Challenge
Across the portfolio, we had a large catalog of existing content that needed ongoing updates to remain accurate, relevant, and competitive in search.
The existing update program depended on freelance writers, but the portfolio did not have enough resources to continue paying for manual updates at the scale required. Leadership believed we might need to abandon the program entirely.
That created a second problem: allowing high-value content to age without updates would cause search visibility to deteriorate and put recurring revenue at risk.
Business Case
I researched the financial impact of ending the update program and found that the portfolio stood to lose approximately $8,000 per month if we allowed those articles to fall out of search results.
Rather than accept the choice between maintaining an expensive manual workflow and abandoning updates entirely, I developed a third option: redesign the program around AI while keeping a human editor responsible for final verification and publication.
My Role
I researched the business risk, designed the AI-first approach, tested the workflow, and established the operating model for producing article updates at substantially greater scale and lower cost.
The initiative focused on using AI where it could create the most leverage—researching what needed to change, drafting new or revised material, and matching the existing author's voice—while keeping human editors responsible for accuracy, relevance, and publication.
AI-First Workflow
The system used AI to perform key research on articles that needed updating to remain competitive in search. It could identify outdated information, determine where new context was needed, and generate replacement or additional sections.
A critical requirement was continuity of voice. The system analyzed the existing article and generated updates intended to match the original author’s tone rather than inserting generic or visibly inconsistent copy.
This shifted the workflow from paying freelancers to research and rewrite every update manually to using AI for the time-intensive first pass, then concentrating human effort where it mattered most.
Human Verification
The workflow was not autonomous publishing. A human editor reviewed the AI-generated changes, verified the facts and current relevance, confirmed that the new material fit the article, and only then published the update.
That human-in-the-loop control allowed the program to increase throughput without treating generated content as inherently trustworthy.
Results
The program transformed a workflow leadership had considered discontinuing into a substantially more scalable operating model, while materially improving the search-performance metrics the update program was designed to protect.
Lasting Impact
The initiative demonstrated that AI could be used as more than a writing shortcut. It became a way to redesign the economics of an established operational program.
By combining AI-assisted research and drafting with human editorial verification, the new model addressed the original resource constraint while increasing output, lowering spend, and improving search visibility.
The broader lesson was operational: when the existing process is no longer financially sustainable, the answer does not always have to be reducing scope. Redesigning the workflow can change the cost structure itself.