The Big Picture: A Genuine Technological Revolution
Redpoint opens with a bold framing: AI is not hype — it's a capital expenditure wave comparable to railroads, electrification, and the internet. Hyperscaler CapEx is projected to hit $765B in 2026 and $779B in 2027, growing at a 36% CAGR from 2024. Critically, they argue this is not a dotcom bubble repeat. Unlike 2000, where fiber utilization was below 3% and investment was debt-funded, today's AI buildout is largely cash-flow funded, over 90% pre-committed, and backed by real revenue — OpenAI and Anthropic are each doing $20B+ ARR.
The AI Product Moment Is Real and Accelerating
The report tracks a remarkable milestone cadence from Jan 2025 through Feb 2026 — DeepSeek R1, Claude Code, Manus Agent, Sora 2, ElevenLabs v3, Waymo's 10-city rollout, and the Claude 4.6/Gemini 3.1 generation of models. AI-enabled software development is exploding: new websites, iOS apps, and GitHub code are all growing 20–50% year-over-year, driven by vibe coding tools like Lovable, Cursor, and Replit.
Disclaimer: Orignal content owned by or sourced from third parties. It does not represent the views of 'Nuggets' platform or it's team. AI is used extensively across this platform including for summaries. Accuracy is not guaranteed, there can be mistakes. Any info or content on this platform is not a financial, legal, or investment advice. Do your own research. Refer for complete disclosures:- Terms of Use · Full Disclaimer
The agent maturity curve is the core thesis for future value creation. We're currently at the "Task Agent" stage (minutes of autonomous runtime), with Workflow Agents (hours) and fully Autonomous AI (days) still ahead — each unlocking a larger addressable market. The total US TAM expands from $0.5T (software spend alone) to potentially $6.1T+ once knowledge worker payroll is included.
The Public SaaS Meltdown
This is where the report gets most forensic. Software stocks are down 20% YTD as of March 2026 — the worst-performing sector, while energy is up 32%. Public SaaS median NTM revenue multiples have collapsed to 4.1x, near all-time lows (vs. a peak of ~22x in 2021). Horizontal SaaS has been hit hardest, down 35% LTM, while Vertical SaaS (-3%) and Infrastructure (+2%) have held up.
The report identifies four forces driving the selloff:
1. Demand Compression — AI captures incremental software spend. 58% of CIOs cite AI feature additions as the #1 driver of increased software spend. Seat-based pricing is under pressure; usage/outcome-based models are expected to dominate.
2. Retention Risk — 54% of CIOs are actively pursuing vendor consolidation, 45% of AI budgets are replacing existing software budgets, and only 3% of CIOs expect AI to lead to more vendors.
3. Terminal Value Compression — In a typical SaaS DCF, 85–95% of enterprise value comes from terminal value. A strong quarter moves less than 5% of that value. Investors aren't debating near-term earnings — they're questioning whether moats survive a decade of AI disruption. An illustrative DCF shows the implied long-term growth rate embedded in current SaaS valuations has fallen from 4.7% to 1.1% in three months — essentially pricing in secular decline.
4. SBC Overhang — After adjusting for stock-based compensation, most SaaS companies show zero real FCF. Companies like Cloudflare, Snowflake, and Atlassian have reported FCF margins of 12–20%, but SBC-adjusted margins turn negative.
The newspaper industry parallel is sobering: newspaper stock prices declined steadily from 2002, but earnings held up for years — until they didn't. Redpoint warns that improving SaaS financials today shouldn't be mistaken for structural resilience.
Why the Selloff Is Differentiated by Layer
Horizontal SaaS (-35%) is the most exposed. These products were "accidentally optimized for replaceability" — built to serve every industry equally, with no proprietary data moats. AI natively solves coordination problems, which is precisely what horizontal tools were built for.
Vertical SaaS (+3%) is relatively protected. Public vertical software companies are deeply embedded systems of record with years of proprietary, industry-specific data. Switching cost is existential. AI is a feature here, not a threat.
Infrastructure (+2%) benefits directly from AI adoption — more AI deployment means more compute, observability, and data spend. The modest decline reflects multiple normalization, not thesis compression.
Debunking the Bear Narratives
Redpoint systematically addresses four "SaaS apocalypse" narratives:
"Vibe coding replaces SaaS" — Concern: LOW. The Slack example illustrates why: buying Slack costs ~$220K/year for 1,000 employees vs. $2M+ to build in-house (and still get an inferior product). Build-vs-buy economics still strongly favor buy for any mature, complex product.
"Startups will displace incumbents" — Concern: HIGH. AI natives are deploying two playbooks — augmenting enterprises (owning the AI intelligence layer atop incumbent systems of record) and attacking SMB/mid-market from below with AI-native UX at lower price points. This is the most credible threat vector.
"Foundation models will do everything" — Concern: MEDIUM. OpenAI launched Frontier and Anthropic launched CoWork, signaling upstack ambitions. But today's models still lack domain logic, collaboration features, integrations, and enterprise UX. Incumbents retain proprietary data advantages. OpenAI has also publicly stated it wants to partner with app companies rather than build every vertical itself.
"SBC is killing everything" — Concern: HIGH. This is real and unambiguous. Software companies continue to grow headcount 5–10% annually while the Mag-7 grows at 1–2%. The labor model needs to change — and AI offers a path to dramatically higher EBITDA margins (projected +100–250% improvement in an AI-optimized business).
The Private Market Opportunity
Despite public market turmoil, private AI investment is booming. VC funding has rebounded sharply, with AI-native deals now comprising ~90% of all new investment by dollar value. The top 20 VC deals represent 44% of all enterprise software VC in 2025, up from just 8% in 2020 — reflecting massive concentration into foundation model labs (OpenAI, Anthropic, xAI) and breakout application companies.
Key private market data points:
Median time to $50B valuation has compressed from 25 years (pre-2000 cohort) to 9 years (post-2000 cohort), with Anthropic at 4 years and xAI at just 1 year
AI-native companies like Cursor ($6.1M ARR/FTE) and Lovable ($3.4M ARR/FTE) are generating revenue per employee 10x higher than traditional SaaS incumbents like Salesforce or Atlassian
The number of companies crossing $10M ARR expanded from 20 (in the first 3.5 years post-ChatGPT) to 50 in just one additional year
Series B/C ARR multiples are at 61x in 2026 YTD vs. public high-growth software at 9.7x — a 528% premium on a raw basis. However, growth-adjusting this actually shows private companies are cheaper — 0.05x growth-adjusted vs. 0.37x for public comparables
Rounds are closing faster (median days between rounds now ~550, down from 700 in early 2024), and valuations are rising at every stage from Seed through Series C.
Redpoint's Investment Framework
Redpoint sees three investment vectors: Applications (expanding into $6T labor TAM), Infrastructure (picks-and-shovels for agent orchestration, security, and observability), and Model Labs (both as portfolio hedges and because new model categories will emerge beyond LLMs).
Their historical analysis of the Internet, Cloud, and Mobile eras shows that the greatest enterprise value is created by companies founded in Years 4–5 of a platform cycle. ChatGPT launched in November 2022 — placing us squarely in the historical sweet spot of 2026–2027.
Redpoint's Honest Unknowns
To their credit, the report concludes with intellectual humility. They cannot predict: which software categories compress fastest, which public market re-rating will arrive, or what foundation model providers do next with their vertical expansion. But their convictions are clear: the $6T labor addressable market is real arithmetic, year 4–5 historically produces durable winners, and the application layer has survived every prior platform shift.
Bottom line: Redpoint's view is that public SaaS is experiencing genuine disruption, not mere sentiment panic, but the private market presents a generational investment window. The companies built in the next 12–24 months — particularly those with vertical data moats, AI-native architectures, and consumption-based pricing — are likely to define the next decade of enterprise software.
(Curated summary using Claude)
"Brookfield's the largest infrastructure owner in the world... We drew a pipeline and we showed all the different components of the payments ecosystem on a pipeline and said it's like a pipe that moves any commodity except what it's moving…