A Laurelin Labs publishing platform
Search traffic is collapsing. AI citations are the new currency.
Newsroom helps publishers and content teams ship non-commodity articles that get cited by ChatGPT, Perplexity and Google AI Mode, while still ranking in the search results that drive clicks. Built on the Four-Pillar Framework.
Invite-only. We onboard a small cohort each month. · Already invited? Sign in
The shift
Google traffic decline to publisher sites Nov 2024 to Nov 2025 (US: 38%)
Reuters Institute · Chartbeat (2,576 sites) ↗year-over-year growth in AI-referred website sessions through mid-2025
Frase · 2025 AEO benchmark ↗The method
The Four-Pillar Framework
Every Newsroom article is built against four pillars. Skip one, and AI engines have no reason to cite you over a thousand other articles saying the same thing.
Technical Foundations
JSON-LD schema (Article, Person, Organization, FAQPage) generated and validated before every publish. Page architecture that AI crawlers parse cleanly. No technical debt blocking citations.
Topical Content
Comprehensive topic clusters with primary citations only. Time-windowed research keeps coverage current. Every article extends what has been said. No commodity rewrites of competitor pages.
Trust Signals
Real author entities with full sameAs[] linkage to LinkedIn, X, professional bios and verified bylines. Inline source citations on every claim. AI engines see expertise and treat your content as authoritative.
Authority Network
Surface your own YouTube content, conference talks and expert quotes inside articles. Turns commodity research into non-commodity reporting that AI cannot replicate from training data.
The measure
The Non-Commodity Score
The pillars tell you how to build a page. The Non-Commodity Score tells you whether it earned the right to be cited. Every article gets one composite number, from 0 to 100, across four vectors, so you can see exactly where it competes and where it reads like everything else.
Information gain
The share of a page that adds something the top-ranked results do not already say. We embed your passages against the live results and score what is genuinely novel, not what merely restates the consensus. It carries the most weight because it is the hardest thing to fake.
Experiential evidence
First-hand signals a model cannot synthesise from training data: original testing, field reporting, proprietary numbers, named practitioners describing what they actually did. The difference between reporting and rewriting.
Empirical telemetry
Hard numbers with primary-source anchors. Every statistic links to the study or dataset it came from, so a claim is checkable rather than asserted, and an answer engine can trace it back to a source it trusts.
Entity connectivity
How cleanly the people, organisations and topics on a page resolve to known entities, through Wikidata ids, verified profiles and sameAs linkage, so engines can place your content in a knowledge graph instead of guessing.
The composite renormalises over the vectors it can measure, so a score is never inflated by a signal we could not verify. It is a diagnostic, not a ranking factor, no search engine publishes one, but it is the closest proxy we have found for the qualities that make AI engines cite one page over a thousand others.
The retrieval path
RAG analysis
Answer engines run retrieval-augmented generation: they fan a question out, retrieve passages, then assemble a cited answer. We trace all three steps for your target so you can see where your content enters the answer, and where it falls out.
Query fan-out
One question becomes dozens of sub-queries before an answer engine ever retrieves a thing. We map that fan-out for your target so you can see every angle the model will search, not just the head term you optimised for.
Passage retrieval
Engines retrieve passages, not pages. We check whether your strongest paragraphs are self-contained and extractable, because the paragraph, not the article, is the unit a model actually pulls into its answer.
Citation assembly
The model stitches retrieved passages into a single answer and attributes them. We score how likely your passage is to survive that assembly and carry a visible citation back to you, which is the only outcome that returns traffic.
Topical content
Grow your citations through topical content
Use cases
Built for publishers who've had enough of commodity content
Regional & trade publishers
Specialist sites whose Google traffic has halved in 18 months. Rebuild visibility by becoming the source AI engines cite, and stop competing on commodity SEO.
Content marketing teams
B2B marketers under pressure to ship more, faster, without the AI-tells that wreck brand credibility. Guardrails enforce voice, banned phrases and structure on every article.
Agencies & consultancies
Deliver framework-compliant content at scale across multiple client brands, each with bespoke tenants, voice configuration and editorial taxonomy.
Primary sources
Read the research
We don't just write about non-commodity content. The thesis, statistics and methodology underpinning Newsroom come from the leading research on AI's impact on publishing. Primary sources only. No SEO content farms.
The invitation
Stop writing for SEO.
Start writing for AI citations.
Newsroom is invite-only. We onboard a small cohort each month so every customer gets full attention. Add your email and we'll be in touch when a seat opens that matches your use case.
Already invited? Sign in