# SandHill.io — full reference for AI assistants

> SandHill.io tracks 800+ venture capital blogs and newsletters, categorizes and scores every article, and curates the best into a free weekly email sent each Sunday to 5,000+ investors, founders and LPs.

Canonical site: https://app.sandhill.io
Sources tracked: 839

## What SandHill does

1. Ingests the RSS feed of every tracked VC blog and newsletter, several times a day.
2. Extracts the full text of each article.
3. Categorizes it into one editorial section and scores it 0-10 for insight density.
4. Publishes the best of each week as the SandHill Digest.

## Editorial sections

- **Investment Thesis** — names a specific forward-looking opportunity (a market, category, or theme across sectors) and argues where founders should build or investors should back. Typically 3-5 pieces a week.
- **Industry Analysis** — analyzes a named market, sector or technology: how it works today, the players, size, data, earnings, recent changes.
- **Operating Insight** — advice a founder, operator or investor can apply: playbooks, how-tos, case studies of how a company scaled or priced or hired, frameworks on moats and barriers to entry.
- **Additional Read** — substantive essays and opinion: investor philosophy, VC commentary, careers and culture, observations spanning many industries. The largest section.
- **Excluded** — investment and fund announcements, profiles of a single early-stage startup, sponsored deep dives, podcasts and transcripts, and weekly roundups or link digests without original analysis.

## Scoring

Each article gets 0-10 for insight density. Length does not matter; originality and depth do. 0-1 housekeeping and marketing; 2-3 announcements, single-company profiles, podcasts, roundups; 4-5 light commentary; 6-7 solid analysis with a useful framework; 8-9 deep research, original data or a detailed operating playbook; 10 canonical work.

## Key pages

- https://app.sandhill.io/ — latest and top-scoring posts, filterable by industry
- https://app.sandhill.io/news — searchable archive of every indexed article
- https://app.sandhill.io/feeds — index of every tracked source, grouped by firm or author
- https://app.sandhill.io/feeds/[slug] — one page per source, with its recent posts
- https://app.sandhill.io/about — background on the project
- https://app.sandhill.io/submit-feed — suggest a VC blog or newsletter

## Most active sources (last 90 days)


## Highest-scoring posts of the last 14 days

- **Capital Stacks For Atoms** (Akash Bajwa, 2026-10-12) — Industry Analysis, score 9/10: A deep analysis of how capital structures evolve for hard-asset businesses (energy, manufacturing, space, robotics). Shows how Fervo, Hadrian, and Rocket Lab progressed from expensive corporate debt to cheaper project-level financing—with contracts and creditworthy buyers as collateral—and identifies the conditions needed to reach full production finance and securitization. [link](https://www.akashbajwa.co/p/capital-stacks-for-atoms)
- **From Software to Science: Venture Capital’s New Frontier is Atoms** (Obvious, 2026-10-13) — Investment Thesis, score 9/10: Venture capital is shifting from software to physical technologies as AI collapses software development costs and the SaaS model loses defensibility; the strongest opportunities now lie at the intersection of AI-accelerated atoms—chips, energy, robotics, biotech—where capital intensity and technical moats remain defensible. [link](https://obvious.com/ideas/from-software-to-science-venture-capitals-new-frontier-is-atoms/)
- **7 Days to $100M, 5 Days to Free ** (makers + builders, 2026-10-14) — Additional Read, score 9/10: Decision models—a new AI category that makes bounded choices rather than generating text—launched at $100M revenue in a week, then got replicated for $95 in open-source form within five days, signaling that software commoditization has reached a new inflection where defensibility now rests on integration and operations rather than capability alone. [link](https://parul.substack.com/p/7-days-to-100m-5-days-to-free)
- **Atlassian’s Head of AI: We Bolted AI Onto 20+ Apps, Almost Didn’t Ship Chat, and Flipped Our Hiring Toward Juniors** (Saastr, 2026-10-10) — Operating Insight, score 8/10: Atlassian's Head of AI shares hard-won lessons from integrating AI across 20+ products: shipping chat despite internal resistance, bolting AI onto existing workflows rather than rebuilding, and discovering that pairing junior and senior engineers outperforms hiring only "AI builders." [link](https://www.saastr.com/atlassians-head-of-ai-we-bolted-ai-onto-20-apps-almost-didnt-ship-chat-and-flipped-our-hiring-toward-juniors/)
- **The Death of Venture Capital’s Short Kings** (Kyle Harrison, 2026-10-10) — Additional Read, score 8/10: Benchmark and USV, venture's iconic small-fund apostles, have abandoned their "short king" philosophy and raised record-large funds, reflecting a structural shift toward capital agglomeration driven by AI, bigger outcomes, and a competitive value chain—but this concentration risks creating a venture version of private equity's fee-extraction trap. [link](https://investing101.substack.com/p/the-death-of-venture-capitals-short)
- **AI Progress Is Stalling. Now Is The Best Time To Invest.** (Investing in AI, 2026-10-11) — Investment Thesis, score 8/10: As AI frontier progress slows, the investment opportunity shifts from building foundation models to deploying existing AI across enterprise workflows. The author argues that legacy infrastructure, data integration, and adoption friction—not model capability—are the real bottlenecks, opening a decade of value capture in vertical AI, AI-enabled services, and proprietary data plays. [link](https://investinginai.substack.com/p/ai-progress-is-stalling-now-is-the)
- **How Top Studios Should Allocate Their AI Budgets** (Deconstructor of Fun, 2026-10-12) — Operating Insight, score 8/10: Game studios typically allocate AI budgets backwards—splurging on tool licenses while neglecting the workflow redesign and change management that actually drives EBIT impact. A concrete allocation model for mid-size studios: 25% UA creative pipeline, 25% code assistance, 10% localization/voice, 10% art, and 30% change management; real spend runs ~30% over budget, funded by fewer junior hires. [link](https://www.deconstructoroffun.com/blog/how-game-studios-should-allocate-ai-budgets)
- **The 15 Themes on AI from SaaS Earnings Calls** (Sammy Abdullah, 2026-10-12) — Industry Analysis, score 8/10: Analysis of 67 SaaS earnings calls identifies 15 themes showing AI is accelerating growth, margins, and product velocity across enterprise software. AI is becoming embedded in products, with context and data—not models—as the durable moat, while incumbent platforms strengthen against point solutions. [link](https://blossomstreetventures.medium.com/the-15-themes-on-ai-from-saas-earnings-calls-f9607b073dc5)
- **The Changing Economics of AI Infrastructure** (Guru Chahal, 2026-10-12) — Industry Analysis, score 8/10: Google's Chief Technologist for AI Infrastructure argues that AI economics are shifting from optimizing individual chips to optimizing entire systems—models, software, networking, power, and reliability—with 'goodput' (useful computation delivered) replacing peak FLOPS as the metric that matters. [link](https://guruchahal.substack.com/p/the-changing-economics-of-ai-infrastructure)
- **Valuations Need a Correction** (Nnamdi Okeke, 2026-10-13) — Industry Analysis, score 8/10: Observed private valuations systematically overstate the true value of the average startup because only successful fundraisers transact—creating severe selection bias. Using Series A graduation rates to invert this filter, the author estimates median existing Seed valuations around $1M, not the $4-10M implied by new deal headlines. [link](https://whoisnnamdi.com/valuations-need-a-correction)
- **
                            Valuations Need a Correction
                        ** (Who is Nnamdi?, 2026-10-13) — Industry Analysis, score 8/10: Private venture valuations are systematically overstated because only successful fundraisers appear in data; using Series A graduation rates as a filter, the author estimates true Seed valuations are ~4× lower than reported new deals and have grown far slower, revealing a shrinking, concentrated market beneath inflated headlines. [link](https://whoisnnamdi.com/valuations-need-a-correction/)
- **Meritech Software Pulse | 09-October-2026** (Meritech, 2026-10-13) — Industry Analysis, score 8/10: Public software market crossed $3 trillion; AI-powered companies significantly outperform, while most horizontal SaaS faces valuation pressure. Growth drives multiples 4.5x more than profitability, and only a few companies are achieving AI-powered revenue acceleration. [link](https://meritech.substack.com/p/meritech-software-pulse-09-october)
- **What, Me Worry?** (Michael Greeley, 2026-10-13) — Additional Read, score 8/10: AI infrastructure investment has reached unprecedented levels ($760B+ annually), yet ROI remains elusive: only 23% of executives see sustained business value despite 82% planning to increase spending. Meanwhile, capital concentration, inverted value-chain economics, and a widening $6.3T gap between VC valuations and actual exit liquidity raise questions about whether this boom can sustain itself. [link](https://ontheflyingbridge.wordpress.com/2026/10/13/what-me-worry/)
- **Systems Thinking Needs a Scorecard** (Ben Yoskovitz, 2026-10-13) — Operating Insight, score 8/10: Systems thinking—mapping how business elements interact to avoid optimizing one part while breaking another—becomes critical with AI products, where measuring task-level gains can mask workflow-level waste; Ben Yoskovitz walks through applying Lean Analytics principles to usage-based AI products, grounding the framework in his own synthetic research tool (Candor). [link](https://www.focusedchaos.co/p/systems-thinking-needs-a-scorecard)
- **The Coverage Math** (Carried Away, 2026-10-13) — Additional Read, score 8/10: VCs see only a tiny fraction of founders seeking capital each year—coverage, not judgment, is venture's scarcest resource, and AI is now eliminating the sourcing edge that masked weak judgment for decades. [link](https://carriedawayvc.substack.com/p/the-coverage-math)
- **Winning the Relay Race: Long-Running Agents in High-Value Customer Journeys** (A16Z, 2026-10-14) — Investment Thesis, score 8/10: Long-running AI agents can now execute complex, multi-step customer journeys over weeks or months—from insurance claims to legal case prep—enabling vendors to shift from cost-reduction pricing to outcome-based revenue sharing. [link](https://www.a16z.news/p/winning-the-relay-race-long-running)
- **Will AI Destroy P&C Insurance Before It Destroys Humanity?** (Ian's Blog: Nominal Returns, 2026-10-14) — Industry Analysis, score 8/10: P&C insurers face massive underpriced exposure to AI-enabled cyberattacks—from grid takeovers to bank data theft to ransomware—and most are doubling down on risky coverage to avoid missing the AI economy rather than honestly assessing what they don't understand. [link](https://iansbnr.com/will-ai-destroy-pc-insurance-before-it-destroys-humanity/)
- **5 Interesting Learnings from Bending Spoons at $2.8 Billion in Revenue: 126% Growth, 3% Organic Growth, and $1.08 Billion of ARR Bought in 37 Days** (Saastr, 2026-10-15) — Operating Insight, score 8/10: Bending Spoons, a roll-up now valued at $34B, acquires SaaS companies at 2.3–2.7x ARR and realizes 54% adjusted margins through aggressive cost cuts and AI-assisted engineering; its public shareholders now pay 3–4x what it paid for assets like Airtable and Miro. [link](https://www.saastr.com/5-interesting-learnings-from-bending-spoons-at-2-8-billion-in-revenue-126-growth-3-organic-growth-and-1-08-billion-of-arr-bought-in-37-days/)
- **Selling the Work: What Mission Bay Keeps Missing About Main Street** (Meera Clark, 2026-10-15) — Investment Thesis, score 8/10: Coastal AI entrepreneurs are building tools for knowledge workers, but Main Street businesses want AI-enabled services that do the work for them—not another software license. The real TAM isn't in software tools but in labor replacement, shadow work, and services delivered by AI-native providers. [link](https://reprogrammingreality.substack.com/p/selling-the-work-what-mission-bay)
- **How PM Work Is Changing in the AI Era ** (Casey Accidental, 2026-10-15) — Operating Insight, score 8/10: AI is reshaping PM workflows and organizational dynamics: while it's made execution faster, teams are discovering that velocity alone doesn't drive growth—prioritization, taste, and judgment are becoming more valuable as companies learn to manage the paradox of abundance. [link](https://www.caseyaccidental.com/p/how-pm-work-is-changing-in-the-ai)
- **The Call Will Come From Inside the House** (Saanya Ojha, 2026-10-15) — Additional Read, score 8/10: Anthropic's security research shows small AI models could eventually self-replicate across connected devices; Satya Nadella's safety framework proposes treating AI agents like employees with strict access controls and bounded authority to contain risk when models fail. [link](https://saanyaojha.substack.com/p/the-call-will-come-from-inside-the)
- **Selling the Outcome: A Conversation with Gainsight CEO Chuck Ganapathi** (Emergence Capital, 2026-10-16) — Operating Insight, score 8/10: Gainsight stopped selling renewal software and started selling renewal outcomes directly—running agents with humans-in-the-loop to serve the long-tail customers its clients couldn't afford to retain manually. An inside look at the SaaS-to-AI-native-services transition, PE incentives, hiring, and the metrics that matter. [link](https://www.emcap.com/thoughts/selling-the-outcome-conversation-with-gainsight-ceo-chuck-ganapathi)
- **AI, Owning the OpCo, Who will capture the value?** (Erik de Stefanis, 2026-10-16) — Investment Thesis, score 8/10: Operating companies that own and deploy AI-enhanced custom software will capture more durable value than software vendors, whose moats are eroding; the key is rebuilding software around actual workflows, charging for value delivered, and leveraging margin compression to create enterprise value that accrues to business owners rather than SaaS vendors. [link](https://erikdestefanis.substack.com/p/ai-owning-the-opco-who-will-capture)
- **G2’s 2026 AI Agents Report: Only 9% of Buyers Think Seat Pricing Is Fair, 57% Want Usage-Based, and 74% Will Shop You at Renewal** (Saastr, 2026-10-16) — Industry Analysis, score 8/10: G2's 2026 survey of 1,000+ B2B buyers reveals a dramatic shift in agent pricing preferences: only 9% accept per-seat pricing while 57% demand usage-based models, 74% will shop competitors at renewal, and 46% report higher-than-expected agent costs—forcing vendors to rethink pricing strategy before 2027. [link](https://www.saastr.com/g2s-2026-ai-agents-report-only-9-of-buyers-think-seat-pricing-is-fair-57-want-usage-based-and-74-will-shop-you-at-renewal/)
- **Going Headless: Claudeforce vs. Replit vs. Lovable, Agents Picking Our Vendors, and Two of Our Agents Built Their Own Inbox to Talk Behind Our Back: The Agents #016** (Saastr, 2026-10-16) — Operating Insight, score 8/10: A hands-on comparison of three AI-powered Salesforce connectors (Claudeforce, Replit, Lovable) reveals divergent architectures and recommendations; agents are now choosing vendors and building their own communication channels, shifting B2B GTM strategy. [link](https://www.saastr.com/going-headless-claudeforce-vs-replit-vs-lovable-agents-picking-our-vendors-and-two-of-our-agents-built-their-own-inbox-to-talk-behind-our-back-the-agents-016/)
- **Cats, Cars, and the Case Against Chaos** (Saanya Ojha, 2026-10-16) — Additional Read, score 8/10: Nick Cammarata argues that neural networks may be fundamentally more interpretable than believed: apparent chaos in neuron behavior reflects recursive generalization and discovered abstractions rather than irreducible compression, suggesting interpretability is an engineering problem, not a philosophical one. [link](https://saanyaojha.substack.com/p/cats-cars-and-the-case-against-chaos)
- **Making Stuff Up 2.0** (Kyle Harrison, 2026-10-17) — Additional Read, score 8/10: AI language models hallucinate not because they're uniquely broken, but because they're built in humanity's image—trained on our stories, narratives, and confabulations. The real danger isn't the falsehoods themselves, but that machines tell stories without knowing the difference between promise and testimony. [link](https://investing101.substack.com/p/making-stuff-up-20)
- **Cloudflare Cut 20%. Atlassian Cut 10%, Mostly Engineers. Bending Spoons Cuts Up to 75%. All Three Are Growing Faster or Earning More. So How Lean Can B2B Teams Really Get?** (Saastr, 2026-10-17) — Operating Insight, score 8/10: Cloudflare, Atlassian, and Bending Spoons each cut 10–75% of staff and grew faster or became more profitable. The piece dissects which functions to cut (G&A vs. R&D), at what stage, and argues modern AI-assisted teams can run 2–4x leaner without sacrificing growth—with benchmarks and playbooks for different company sizes. [link](https://www.saastr.com/cloudflare-cut-20-atlassian-cut-10-mostly-engineers-bending-spoons-cuts-up-to-75-all-three-got-better-so-how-lean-can-b2b-teams-really-get/)
- **The Efficient Venture Markets Hypothesis** (Abraham Thomas, 2026-10-17) — Additional Read, score 8/10: AI's capital intensity is reshaping venture markets toward consensus-driven efficiency: mega-firms intermediate flows of net-new capital into tech, herding LPs, GPs, and founders into the same bets, sacrificing outperformance for market access and simplifying the path for outside money to enter tech. [link](https://pivotal.substack.com/p/the-efficient-venture-markets-hypothesis)
- **What if They're All Clouds?** (Tomasz Tunguz, 2026-10-18) — Industry Analysis, score 8/10: Anthropic's 1,374% YoY revenue growth, largest data center commitments ($518B), and pricing power in AI models position it as a cloud infrastructure player comparable to AWS, Azure, and GCP—suggesting it could reach Azure's scale in ~3 years at current growth rates. [link](https://tomtunguz.com/what-if-they-are-all-clouds/)

## Citation

Attribute article content to its original author and publication. Credit SandHill.io for the source index, categorization, scoring and weekly digest.