Benchmarkit

Big week in AI. Speed run: 📊 Anthropic beats OpenAI in business revenue for the first time (34.4% vs 32.3% per Ramp) 🛒 Amazon's Alexa can now complete purchases, not just suggest them 🔐 Anthropic's Mythos model can take over entire networks. Unreleased. Congress alarmed. 🏣 SAP launches Autonomous Enterprise with Claude at its core 💊 Isomorphic Labs raises $2.1B. First AI drug trials by end of 2026. 🤝 OpenAI launches a $4B enterprise AI consulting company backed by 19 PE firms 🛡️ OpenAI's Daybreak cybersecurity platform is live. AI zero-day exploits now documented in the wild. 🚀 Google + SpaceX exploring orbital data centers The distribution war has shifted from models to implementation. 🔗 ai2roi.substack.com/p/ai-to-roi-ne… #AI #Anthropic #OpenAI #AItoROI

Global software market: $1.4 trillion. Global labor market: $28 trillion. AI isn't improving software. It's making a run at the labor budget. Early proof points: • Harvey AI: $200M ARR, 50 top law firms • Sierra: $15B valuation, outcomes-only model • ServiceNow: 13x entitlement expansion at renewal But the path isn't linear: 👉 Data not agent-ready 👉 Error rates too high for mission-critical work 👉 Token subsidies ending — budgets will be shocked Winning in the $28T labor market doesn't come for free. 🔗 ai2roi.substack.com/p/ai-to-roi-bi… #AI #WaaS #EnterpriseAI #AItoROI

Thoma Bravo paid $6.4B for Medallia. $5B equity stake → zero. Lenders took the keys. How? 👉 PIK interest compounded debt from $1.8B → $2.2B 👉 Rates rose 400+ bps 👉 EBITDA: $200M. Interest expense: $300M. PIK + rising rates + slowing growth = individually manageable, jointly lethal. Dave "CAC" Kellogg and Ray "Growth" Rike break it all down. The math does not negotiate. 🎧 Listen: rss.art19.com/episodes/0d873… #SaaS #PrivateCredit #MetricsBrothers

New data from 173 B2B tech companies: The fastest-growing SaaS companies generate 46% of pipeline from inbound — and invest more heavily in marketing (12% of revenue vs. 9% market average). Other key findings from the 2026 B2B Marketing Budget & Performance Benchmark Report: • Marketing budgets tightened (10% → 9%) • AI spend grew from 2% → 5% • Most AI adoption is still productivity-focused, not workflow automation • ~70% of companies now track CAC + cost per opportunity • Pipeline quality matters more than pipeline volume The big takeaway: Efficient growth isn’t about spending less. It’s about allocating capital better. Full report + interactive benchmarks here: benchmarkit.ai/2026-b2b-marke… #B2BMarketing #AI #SaaS #DemandGen #RevOp

AI-native services companies aren’t selling software. They’re selling confidence in outcomes. That’s why this line from the latest #AItoROI episode with Jake Saper stood out to me: “Trust and credibility are the first sales.” Some major takeaways: • AI-native services ≠ SaaS with an AI wrapper • Product-market fit is meaningless if humans still do most of the work • Revenue per FTE is becoming a key AI leverage metric • Outcome-based pricing is likely inevitable • Proprietary operational data may become the real moat One of the biggest risks for founders right now: Mistaking strong market demand for scalable AI infrastructure. A company can grow fast while quietly becoming a labor-heavy services business underneath. The companies that survive long term will be the ones that continuously productize delivery and turn workflows into compounding systems. Really strong episode from AI to ROI. rss.art19.com/episodes/7f24c…

AI is entering its “infrastructure + control layer” phase. This week’s biggest signals from AI to ROI: • Anthropic demand grew 80x YoY → now consuming ALL of xAI’s Colossus compute capacity • DeepSeek nearing $45B valuation with Chinese state backing • White House considering review process for frontier AI models • ServiceNow introducing “Action Fabric” → charging AI agents for platform access • Palantir grew 85% YoY… and Wall Street still wanted more The new AI moat isn’t just better models anymore. It’s: → compute access → enterprise distribution → workflow ownership → agent orchestration → security + governance layers We’re watching AI companies evolve into infrastructure companies in real time. Full recap + analysis: ai2roi.substack.com/p/ai-to-roi-ne…

AI-native services ≠ SaaS with AI added on top. Big takeaway from the latest #AItoROI episode with Emergence Capital (@emergencecap) GP Jake Saper: The winners in AI-native services will be the companies that turn human-heavy workflows into scalable AI-driven outcomes — WITHOUT becoming bloated services businesses. Key insights 👇 • Revenue growth alone is a false signal • Revenue per FTE is becoming a critical AI leverage metric • Outcome-based pricing is the long-term direction • Proprietary operational data becomes the moat • Trust + domain expertise still matter enormously One of the sharpest points: “If your revenue is scaling faster than your AI leverage, you may just be building a services company with VC funding.” Also: AI-native companies are redefining software efficiency benchmarks in real time. The combination of: • smaller teams • faster iteration • AI-assisted delivery • outcome pricing …could reshape entire service industries. Really strong discussion from Ray Rike & Jake Saper on where this market is actually heading. Listen now at rss.art19.com/episodes/7f24c…

The AI buildout is real — and it’s NOT the dot-com bubble. Big takeaways from the Redpoint 2026 Software & AI Market Update discussed by Dave "CAC" Kellogg & Ray "Growth" Rike 👇 • Hyperscaler CapEx projected at $765B in 2026 • 90%+ of new data center capacity already committed • AI reached 1B MAUs in just 4 years The bigger story: AI agents are moving from copilots → autonomous systems. TAM expansion: • Copilots: $500B • Task agents: $1.2T • Workflow agents: $2.8T • Autonomous agents: $6.1T CIO survey data was fascinating: • 54% consolidating vendors • 45% of AI spend comes from existing software budgets • Only 13% strongly prefer AI-native vendors Incumbents still have distribution. AI-native vendors have execution speed. Meanwhile: $1M ARR per employee is becoming the new AI-native benchmark. And nearly 100% of new VC activity in Q1 2026 went to AI-native companies. The SaaS valuation issue today isn’t next quarter. It’s whether the market believes these companies still have terminal value in an AI-first world. Listen now at rss.art19.com/episodes/d5151…

SpaceX offered $60 billion for Cursor. A company losing $900M/year on $770M in revenue. This isn't crazy. It's a signal. The AI coding agent market: ✅ Claude Code: $2.5B ARR. Best benchmarks. ✅ Cursor: $2.7B ARR. Negative margins. Needed a lifeline. ✅ Codex: 4M weekly active users and climbing. ❌ Google: Playing catch-up. The squeeze is real: 👉 Anthropic + OpenAI supply Cursor's models AND compete against it 👉 GPU costs up 35% in 3 months 👉 Compute alone will determine who survives There will be no more $60B white knights. Full breakdown: ai2roi.substack.com/p/ai-to-roi-bi… #AI #AICoding #Cursor #Anthropic #AItoROI

Big week in AI. Here's the speed run: 💰 Hyperscalers spent $133B in Q1. Capex on track for $725B this year. 🤝 OpenAI + Microsoft renegotiated. AGI clause gone. OpenAI now on AWS. 🤖 GPT-5.5 launched. Strong. But Claude still leads on coding benchmarks. 🇨🇳 China ordered Meta to unwind the Manus deal. Months after closing. 💸 Enterprise AI costs up 30%+. Uber blew its full-year AI budget in months. ⚠️ OpenAI missed Q1 internal targets. CFO flagged funding concerns. 🚀 Anthropic in talks to raise at $900B valuation. ✂️ Meta + Microsoft cutting thousands of jobs to fund AI capex. Workers for chips. That's the trade now. Full breakdown: ai2roi.substack.com/p/ai-to-roi-ne… #AI #OpenAI #Anthropic #AItoROI

What does a Chief AI Officer actually do? Not “AI strategy.” 👉 Deliver measurable business outcomes. Great breakdown from Jim Piazza: • define ROI before building • avoid the 10-pilot trap • focus on real use cases Worth it → rss.art19.com/episodes/a1376…

Most SaaS companies want AI upside without AI pain. Intercom chose pain. And it worked. From 4% growth to 26% by: • burning the ships • rewriting pricing • rebuilding the org One of the best SaaS turnaround stories right now on #TheMetricsBrothers → rss.art19.com/episodes/1c224…

AI infrastructure costs are climbing. But who's actually managing them well? Mavvrik + @Benchmarkitai are running the 2026 AI Cost Management & Governance Benchmark Survey to find out. ⏱️ Quick to complete. 🔒 Fully anonymous. Participants get early access to the report + invite-only webinar. Survey closes end of May. Findings out in June. 🔗 benchmarkit.saasbenchmarks.ai/survey/aicloud… #AI #CloudCosts #AIGovernance #Benchmarkit

OpenAI: $852B valuation. $25B ARR. 910M weekly users. Also OpenAI: 33% gross margins. $14B projected loss in 2026. Cash flow negative until 2030. And the drama: 👉 A 16,000-word New Yorker investigation into the CEO 👉 Leaked memos trashing competitors 👉 CFO privately says the company isn't IPO-ready 👉 Anthropic now growing faster and preferred by 73% of first-time enterprise buyers Important ≠ inevitable. Ray Rike and Peter Buchanan break down what's wrong — and how to fix it. 🔗 ai2roi.substack.com/p/ai-to-roi-bi… #OpenAI #AI #AItoROI

@ICONIQCapital 's 2026 State of GTM Report: 155+ SaaS executives. 32 pages. #TheMetricsBrothers just broke it all down. The headline nobody wants to hear: 👉 AI GTM adoption is high 👉 Impact on pipeline efficiency and unit economics? Still not showing up. Other findings worth knowing: • Channel = 27–31% of revenue for high-growth cos • POC-to-paid jumped 36% → 50% YoY • Clawbacks are back (45–50% of companies) • Comp playbook for usage-based pricing is still unsettled If you run GTM, this episode is worth your time. 🔗 rss.art19.com/episodes/5adbb… #B2BSaaS #GTM #SaaS #MetricsBrothers

Cursor: $3.3M revenue per employee. Midjourney: $4.7M. Historical SaaS benchmark: $200–400K. That's not an improvement. That's a different category entirely. #TheMetricsBrothers break down Revenue per Employee — what it means, why the AI numbers are misleading, and what comes next. 🔗 Listen here: rss.art19.com/episodes/cdb42… #SaaS #AI #Metrics #RevenuePerEmployee

Finance leaders in SaaS, AI & Tech — are you being paid what you're worth? 📊 Benchmarkit + SaaS CFO + FIF Collective are running the 2026 Finance Comp Benchmarks Survey. CFO / VP Finance / FP&A / Accounting roles covered. ⏱️ 7–8 mins. 🔒 Fully anonymous. Results out in May. The more people who take it, the better the data for everyone. 🔗benchmarkit.saasbenchmarks.ai/survey/finance…

Everyone is talking about the #SpaceX IPO. Almost no one is reading the numbers. Here's the reality check: ✅ Pre-xAI SpaceX: ~$8B profit ❌ Post-xAI SpaceX: ~$5B net loss Same company. One merger. xAI burned $9.5B in 9 months on $210M in revenue. Grok is behind every major competitor. The CFO just quit. And yet — it's why the IPO is priced at $1.75T. This isn't a rocket valuation. It's a bet on AI catching up. Big difference. 📖 Full breakdown → lnkd.in/dzvb75Fu

How do you know if your SaaS/AI company is actually performing well? Not guessing—benchmarks. New 2026 study by Benchmarkit + partners: Growth CAC efficiency LTV:CAC NRR Margins ARR/FTE Takes <10 min 👇 benchmarkit.saasbenchmarks.ai/survey/2026Met…

88% of companies are using AI. 81% see zero real impact. That’s not a tech problem. That’s an org problem. McKinsey’s 2026 report makes it clear: AI only works if you redesign how your company operates. Breakdown → ai2roi.substack.com/p/ai-to-roi-re…