❓ Problem

Every company has a blog, and increasingly every company's blog reads like it was written by the same intern who never sleeps. That's because, in a sense, it was. Roughly half of all new articles published online are now primarily AI-generated, and the share has held there for over a year. Another "Five Trends Transforming Your Industry" lands in a feed already drowning in them.

The results show it. In Orbit Media's 2026 survey of 1,042 content marketers, just 13.9% said their blog delivers strong results, the lowest figure in the study's 12-year history. Meanwhile 92.4% now use AI to write, and that adoption correlates with nothing. Everyone got faster. Nobody got better.

Here's the kicker. The one thing AI can't prompt into existence is data nobody else has, and most American B2B companies are sitting on mountains of it: what customers buy, which tools are spreading, where budgets are moving, which behaviours are quietly taking off. Original research works. Nine in 10 companies that publish it say it's successful, yet fewer than half ever do. The stories are sitting in the warehouse. There's just no journalist on payroll to go and find them.

✅ Solution

DataPress is an AI content engine that turns a company's own data into original research, reports and posts. Content with receipts.

• Connect approved data sources. Plug in the warehouse (Snowflake, BigQuery, Postgres) or product analytics, with the security and legal team choosing exactly which tables and fields are in bounds. Everything is analysed in aggregate, never at the customer level.
• Get a weekly story feed. The engine scans for patterns worth writing about: a category growing 3x faster than the rest, a behaviour flipping in one state, a tool quietly overtaking an incumbent. Each pitch arrives like a reporter's tip, with the headline number, a supporting chart and a plain-English note on how it was calculated.
• Pick an angle, generate a draft. The content team reviews the evidence, chooses the framing (press release, LinkedIn post, blog, quarterly report) and DataPress drafts it in house style. Every claim links back to the query that produced it, so nothing ships that can't be traced.
• Approve, publish, repeat. Nothing goes live without sign-off from a named reviewer. Recurring findings can be turned into a standing index, a monthly number the company owns and the press starts quoting.

📊 Key Numbers

Market size

• The US is the biggest content marketing market in the world, estimated at around $160 billion in 2025 out of a global market worth roughly $533 billion. Analysts draw the boundary very differently, so treat these as directional, but either way it is an enormous pool of spend chasing a shrinking supply of attention.
• The US is home to around 17,000 SaaS companies, roughly 60% of the global total. Nearly all sit on usage or transaction data, and most publish content every week. Add American fintechs, marketplaces, payroll, logistics and HR platforms and the serviceable market runs well into the tens of thousands.
• The alternative is expensive. A dedicated data journalist runs around $80,000 a year in salary alone, before benefits, tooling and an analyst to pull the numbers. A single branded research report from a specialist firm starts at about $5,500, and full agency projects run well into five figures.

ARR potential

• Bottom-up: 1,000 US customers at an average of $1,500/mo is $18M ARR. Layer on 200 flagship reports a year through the editorial service at ~$15k each and that's another $3M.
• Top-down: winning 10% of America's SaaS companies alone at ~$18k a year is a ~$31M ARR business, before touching fintech, marketplaces or the enterprise.