**Piece being analyzed:** A LinkedIn post by a VC arguing for remote-first companies based on portfolio data, with a recommendation to founders. **Claim inventory:** 1. Remote-first companies grow 30% faster than in-office companies. 2. The data showing this comes from the author's VC portfolio. 3. The growth rate gap is sustained across multiple years. 4. The mechanism is faster hiring and lower payroll cost. 5. Recommendation: founders should default to remote-first. **Per-claim trace:** **Claim 1 (30% faster growth):** - Support in the text: A single sentence reference to portfolio data. - Support type: argument from authority + unverified summary statistic. - Strength: weak. - Verifiable: not as presented. The portfolio companies aren't named; the methodology for "growth" isn't given; the comparison base (which in-office companies?) isn't defined. - Notes: 30% is a large effect. Claims of large effects with no methodology disclosed should be treated as suggestive, not as evidence. **Claim 2 (data is from author's portfolio):** - Support in the text: stated by the author. - Support type: direct attribution. - Strength: moderate as a sourcing claim, but inherits the weakness of claim 1 — the portfolio data itself isn't shown. - Verifiable: in principle, by asking the author. In practice, unlikely to be shared. - Notes: portfolio data has known selection problems. The portfolio is a sample selected for venture-backable characteristics; whatever pattern emerges may not generalize to the broader founder population. **Claim 3 (sustained across years):** - Support in the text: asserted in one sentence; no time series shown. - Support type: unsupported assertion. - Strength: absent. - Verifiable: no. **Claim 4 (mechanism: hiring speed and payroll cost):** - Support in the text: brief explanation in one paragraph; no quantification. - Support type: argument by plausibility. - Strength: weak. The mechanism is plausible but other mechanisms (e.g., remote-first companies systematically choose specific markets) are equally plausible and not addressed. - Verifiable: in principle, yes, but the piece offers no path. **Claim 5 (founders should default to remote-first):** - Support in the text: rests on claims 1-4. - Support type: recommendation derived from preceding claims. - Strength: as strong as the weakest claim it depends on, which is weak. - Verifiable: not a verifiable factual claim; it's a recommendation. But the basis for the recommendation is weakly supported. **Patterns across the piece:** The piece is front-loaded with a strong-sounding headline statistic, and the supporting claims thin out as the piece progresses. The mechanism explanation does most of the persuasive work despite being qualitative. The recommendation is presented as if it follows from rigorous data, but the data is essentially one undocumented stat. **Weakest claims:** - Claim 1 (the 30% number). Carries the entire weight of the piece; supported by one undocumented reference. - Claim 3 (sustained across years). Unsupported assertion that materially affects whether the headline number reflects a stable pattern or a one-time blip. **Strongest claims:** - Claim 2, narrowly — the author is in a position to know what's in their portfolio. The piece earns the right to say "in our portfolio, we see X," though it doesn't earn the right to generalize.