🕶️ Glean's $300M trick: use less AI

The enterprise AI startup pulling away right now is not the one using the most tokens. It's the one helping customers burn fewer of them.

That shift, from capability theater to cost control, explains how Glean just tripled to $300M while fending off Google, Microsoft, and OpenAI in the same category.

Elsewhere, the weekend's moves skew industrial: SpaceX cutting its IPO ambition, Nvidia betting billions on photonics, and Autodesk buying its way from design into operations. The through line isn't hype. It's margin discipline dressed up as product strategy.

A year ago, most marketers weren't thinking about AI search. Now it's one of the fastest moving channels in the industry and nobody has a playbook yet.

So we built one. We surveyed hundreds of marketers to find out how they're approaching answer engine optimization, where they're investing, what's actually working, and what isn't.

The result is the 2026 State of AEO Report. Real data. Real strategies. A clear picture of where AI search is headed and how to get ahead of it.

Download the free report

The enterprise AI startup pulling away right now is not the one using the most tokens. It is the one helping customers burn fewer of them.

The proof is Glean's $300 million ARR milestone, up threefold from the $100 million mark it hit just 15 months ago. The seven-year-old company was last valued at $7.2 billion after a $150 million Series F last June. One important caveat: the number is not pure traditional ARR, because part of it comes from annualized consumption revenue. That distinction matters as more AI companies move to hybrid pricing.

The speed here is the real signal. Seven years ago, Arvind Jain started Glean as an enterprise search company. Fifteen months ago, it crossed $100 million. Now it is past $300 million even as Google, Microsoft, OpenAI, Anthropic, Salesforce, and Atlassian all push into the same market.

Meanwhile, on the incumbent side… the giants have distribution and model scale, but Glean is betting that enterprise context matters more. Its core pitch is a context graph that connects to internal company systems, learns how those systems relate, and makes AI search more precise. Jain's sharper claim is the strategic one: plugging AI into Glean can materially reduce token consumption, which turns cost savings into product differentiation.

The less obvious shift is in how enterprise buyers will evaluate AI software. Capability alone is no longer enough. If two vendors can answer the same question, the one that uses less compute will increasingly win procurement. That pushes the whole category toward deeper connectors, better retrieval, and pricing models that reflect actual model usage instead of pretending everything is still seat-based SaaS.

Track token savings like revenue: If your product touches LLM workflows, quantify the compute you eliminate. Buyers now need a cost story, not just a demo.

Build connectors like they are your moat: Glean's edge is not branding. It is enterprise-specific context. Low-latency, high-coverage integrations are becoming defensibility.

Be precise about ARR: If you use consumption or hybrid pricing, separate subscription ARR from annualized usage run rate. Sophisticated buyers and investors will care.

🎯 Lucra Sports lands $20M Series B from ARK, defying VC AI-only mindset with creative pitching and solid growth.

What changed ARK’s mind wasn’t a buzzword; it was a clean story tying AI mania to Lucra’s actual retention metrics and white-label economics, which TechCrunch notes helped the fund overcome its prior esports loss on Skillz in this breakdown.

🤖 Seattle’s Picnic shuts down after burning through $50M, sells pizza-robot assets to a mystery buyer.

The shutdown caps a 9-year run propped up by venture, strategic, and government money, including a $5M COVID-era restaurant revitalization grant that never translated into sustainable demand. Filings show Picnic was already pivoting on pricing and go-to-market while the robotics hype cycle was peaking again.

📊 Startups boosting productivity by 40% with performance management systems, turning chaos into clarity.

Leaders didn’t get there with annual reviews; they built lightweight, iterative systems tied to weekly feedback, clear decision rights, and visible rewards, raising the question of when a “simple” system quietly becomes a company’s real operating model.

🤖 ClickUp lays off 22% of staff, replaces roles with 3,000 AI agents and eyes “million-dollar” salary bands for top AI users.

The bet isn’t just fewer humans, it’s a new org chart where “agent orchestrators” sit on top and everyone else becomes a cost center to be automated away. If ClickUp can turn its internal AI metrics into a product, as hinted to TechCrunch, the real leverage may come from selling that playbook to other companies.