Editorial

Methodology

How we choose what to cover, how we process each story, and what controls we apply before publishing.

Last updated: 2026-04-25

1. Source monitoring (RSS)
Every hour we poll the RSS feeds of about 25 pre-approved publishers in real estate, finance, and AI. We filter by relevant keywords and discard sponsored content, opinion without attribution, and duplicates.
2. Extraction and cleanup
For each selected story, we extract the article body from the original source (no scripts or ads), preserving facts, figures, and direct quotes. We keep the link, original author, and publication date.
3. AI-assisted rewriting
Content is rewritten by a language model under a strict editorial prompt: neutral journalistic tone, EN↔ES translation, structure with subheadings, verifiable figures preserved verbatim, and mandatory attribution to the original source. The AI does not invent figures or quotes.
4. Automated validation
Each output passes rules that check: minimum length (950+ words per language), tag count, absence of promotional language, presence of source, empty paragraph counts, and ES/EN consistency. Articles that fail are discarded.
5. Gradual publication
We cap publication at a maximum of 2 articles per hour with 60-minute spacing. This avoids content avalanches and lets each story have visibility on the home and newsletter. The queue is processed at peak reading windows.
6. Multi-channel distribution
Once published, the article is included in the sitemap, RSS feed, newsletter (in morning, afternoon, and evening windows), and Twitter (in ET traffic windows). Links always point back to the original article as source.

Sources we monitor

Bloomberg, Reuters, CNBC, The Wall Street Journal, Financial Times, HousingWire, Realtor.com News, Inman, The Real Deal, Curbed, Architectural Digest, Mansion Global, MIT Technology Review, TechCrunch, VentureBeat, Wired, The Verge, Ars Technica, MarketWatch, Yahoo Finance, Investopedia, Forbes Real Estate, REIT.com, Nareit, FRED (Federal Reserve Bank of St. Louis).

← Editorial Policy Corrections Policy →

Methodology

How we choose what to cover, how we process each story, and what controls we apply before publishing.

Last updated: 2026-04-25

1. Source monitoring (RSS)

Every hour we poll the RSS feeds of about 25 pre-approved publishers in real estate, finance, and AI. We filter by relevant keywords and discard sponsored content, opinion without attribution, and duplicates.

2. Extraction and cleanup

For each selected story, we extract the article body from the original source (no scripts or ads), preserving facts, figures, and direct quotes. We keep the link, original author, and publication date.

3. AI-assisted rewriting

Content is rewritten by a language model under a strict editorial prompt: neutral journalistic tone, EN↔ES translation, structure with subheadings, verifiable figures preserved verbatim, and mandatory attribution to the original source. The AI does not invent figures or quotes.

4. Automated validation

Each output passes rules that check: minimum length (950+ words per language), tag count, absence of promotional language, presence of source, empty paragraph counts, and ES/EN consistency. Articles that fail are discarded.

5. Gradual publication

We cap publication at a maximum of 2 articles per hour with 60-minute spacing. This avoids content avalanches and lets each story have visibility on the home and newsletter. The queue is processed at peak reading windows.

6. Multi-channel distribution

Once published, the article is included in the sitemap, RSS feed, newsletter (in morning, afternoon, and evening windows), and Twitter (in ET traffic windows). Links always point back to the original article as source.