Marcus B. is the Editor-in-Chief of Herald Digital, an independent news organization covering technology and startup culture. Unlike aggregation-focused publishers, Herald Digital built its reputation on original reporting — analysis pieces, trend coverage, and deep-dive features that their readers can't find elsewhere.
When Marcus first evaluated Newsmill, he wasn't looking for a content aggregation tool. He was looking for a way to augment his small editorial team's research capacity. The Originals pipeline gave him exactly that.
The Challenge
Herald Digital's editorial model is straightforward but resource-intensive. The team identifies a story angle, researches it across multiple sources, drafts an article, fact-checks, edits, and publishes. A single well-researched piece takes 4–8 hours of editorial time.
With a three-person editorial team, Herald Digital published 8–10 original articles per week. Their audience engagement was strong — readers valued the depth and perspective — but the production pace limited their ability to cover emerging stories quickly.
The tension was between depth and speed. Herald Digital couldn't compete on volume with aggregation-heavy publishers, but their audience expected them to weigh in on breaking stories within hours, not days. Missing a major story because the team was deep in a long-form piece happened more often than Marcus wanted to admit.
The team experimented with generic AI writing tools, but the output was unusable. ChatGPT-style tools produced surface-level content that lacked the analytical depth Herald Digital's readers expected. The articles read like summaries of summaries — factually thin and editorially empty.
The Solution
Herald Digital adopted Newsmill's Originals pipeline — not to replace their editorial process, but to accelerate it. The workflow changed from "research everything from scratch" to "start from a research-backed draft and add editorial value."
Research-Backed Drafts
Marcus or a team member defines a brief: the topic, angle, key questions to address, and target audience. Newsmill's Originals pipeline researches the topic across its source network, identifies relevant data points and quotes, and produces a structured draft that addresses the brief.
The draft isn't a final article — it's a starting point that would have taken 2–3 hours of research to assemble manually. The editorial team spends their time adding perspective, challenging assertions, and refining the argument rather than gathering basic facts.
Quality Controls
Every Originals draft passes through the same quality pipeline as aggregated content. Humanization ensures the output reads naturally. Deduplication checks verify the draft doesn't overlap with recently published Herald Digital content. Source attribution is preserved so editors can verify claims against original reporting.
The editorial team maintains full control over the final product. No article publishes without human review. Newsmill handles the research grunt work; Herald Digital editors handle the judgment, voice, and editorial standards.
Publishing Workflow
Approved articles are pushed to Herald Digital's static site via Newsmill's Markdown export and publishing pipeline. The integration fits their existing Git-based editorial workflow — drafts appear as files that editors review, revise, and merge.
The Results
After four months using the Originals pipeline, Herald Digital saw meaningful changes in their production capacity and coverage:
- 12–15 original articles per week — up from 8–10, a 50% increase
- Same editorial team — no additional hires required
- Faster breaking news coverage — research-backed drafts available within hours of a story breaking
- Maintained editorial quality — reader engagement metrics (time on page, return visits) stayed flat, indicating no quality drop despite higher volume
The most significant change was qualitative. Marcus described it as "giving each editor an extra research assistant." Stories that previously required a full day of research could be started from a structured draft, allowing the team to cover more ground without cutting corners on depth.
Key Takeaways
Herald Digital's approach to AI-assisted original content differs from typical aggregation workflows, but the underlying principles are consistent:
-
AI as research accelerator, not writer replacement. Herald Digital doesn't publish AI output directly. They use it to compress the research phase, then apply human editorial judgment to the draft. The AI handles the information gathering; humans handle the insight.
-
The Originals pipeline serves a different need than the Feed pipeline. Aggregation automates volume. Originals automate research. Herald Digital uses Originals because their competitive advantage is analysis, not speed-to-publish on commodity news.
-
Quality controls matter more for original content. When you're publishing under your own byline (not aggregating from attributed sources), detection resistance and factual accuracy are non-negotiable. Newsmill's humanization layer and source attribution features were prerequisites for Herald Digital's adoption.
If your editorial team spends more time researching than writing, explore the Originals pipeline or reach out to discuss how AI research assistance could fit your editorial workflow.