Is the article about them, or just naming them?
When someone tells you a company is all over the news, open the articles. In June 2026 Lovable, the AI app builder, seemed to be everywhere. We, meaning me and my AI team, took twelve articles from that month that name it and read each one in full, looking for the sentence where Lovable shows up.
Four were about Lovable itself. In the other eight, Lovable was a name in someone else’s news: three of them Anthropic’s model launch. Here is each one, with the sentence where Lovable shows up:
| Article (source, date) | Where Lovable shows up | Lovable is |
|---|---|---|
| “Lovable Expands Google Cloud Partnership in Major Push …” (CXO Digital Pulse, 4 Jun) | “Lovable has entered into an expanded multiyear partnership with Google Cloud” | the story |
| “Lovable Expands Google Cloud Partnership To Scale AI …” (The AI Software Report, 25 Jun) | the same deal: “Lovable is moving its AI-powered software creation platform onto Google Cloud” | the story |
| “Making the app building experience Lovable” (Perspectives, 27 Jun) | “the biggest player in the prompt-to-app market today: Lovable”, set against Microsoft | the story |
| “Connect Google Analytics 4 (GA4) to Lovable” (Windsor.ai, 19 Jun) | a how-to for Windsor.ai’s own connector | the story, told by another company selling its product |
| “Anthropic launches Claude Sonnet 5 as a cheaper way to run agents” (TechCrunch, 30 Jun) | Lovable’s co-founder, quoted praising Anthropic’s new model | passing: a customer quoted in Anthropic’s launch |
| The same TechCrunch story, republished (Business Ghana, 1 Jul) | the same quote | passing |
| “Claude Sonnet 5: near-Opus power at a Sonnet price” (go-to-agency, 30 Jun) | one of the customers Anthropic quotes | passing |
| “Anthropic releases Claude Fable” (AI with Ara, 10 Jun) | one item in a daily AI roundup: “Lovable Crosses $500M in Annualized Revenue” | listed: one item among many |
| “a16z and top VCs increase seed round investments …” (KuCoin, 26 Jun) | its revenue growth, as an example of how fast AI startups grow | passing |
| “The VC that turned a $750M Anthropic deal into a $14B stake …” (Tech Funding News, 24 Jun) | one of the companies the fund backed; the article is about the fund | passing |
| “Same round, different price tag” (PitchBook newsletter, 22 Jun) | an example of a funding round paid in two parts | passing |
| “Is AI talking about your product in the right way?” (Growth Unhinged, 21 Jun) | its pricing FAQ, as an example of a marketing tactic | passing |
Twelve articles. Four about the company, and on a strict read three, because the how-to is another company selling its product.
How do you tell if an article is about a company or only mentions it?
More than half of what looks like coverage of a company is a passing mention. We sorted 590 articles that named 95 startups in the year before each one raised money. In 329 of them (56 in 100) the startup was only named in passing. In 186 (32 in 100) it was the story. In 43 (7 in 100) it was one name in a list, and in 32 (5 in 100) one of a few companies the article discussed.
We separate being the story from being named. In Ins7ghts, our research platform (a knowledge graph of the data and AI news we read every day), that sorting comes before any count.
The check, step by step
- Find the company’s name in the article and read the sentence around it.
- Ask: is the article about something the company did, a launch, a deal, a hire, a customer, a result? Then it is the story.
- If not, decide which of the other three it is: one of a few companies the article discusses, one name in a list, or named in passing.
- Count only the articles where it is the story, plus the few where it is discussed if you want a wider read.
- Set aside job ads and the company’s own pages: that is the company talking, not coverage.
The company’s own voice is the next trap: about half of the company stories we read are the company talking about itself. That test gets its own article on 15 December.
The kit has the prompts and the checklist. It comes by mail once you confirm your email.
Small print. 590 articles that named 95 startups in the year before their funding round (2025 and 2026), at most 12 articles per startup. An AI model sorted each article; we read 18 of its labels, picked at random, and they were right, but 18 is a spot check, not a measurement. All 95 startups did raise money; we did not sort the coverage of startups that never raised, so being the story says nothing about what comes next. The twelve Lovable articles were read in full on 8 November 2026.