Top 10 in Tech - What to know for Week ending November 28, 2025

​1. SaaS METRIC OF THE WEEK: GES: Growth Endurance Score is a metric that assesses a company's ability to sustain growth over time. GES measures this efficiency by factoring in both net retention and customer acquisition efficiency. A high GES correlates with long-term business health and resilience. This score provides valuable insights for businesses aiming for consistent, sustainable growth. Bessemer has drilled deeper into it and plotted ARR growth lost YoY, and found that the decay is fairly predictable at 30%. That's a benchmark - in other words, you should expect next year's growth rate to be 70% of the current year, as the stakes get higher.


2. AI BENCHMARKS: HubSpot benchmarked 500 startups' AI-GTM setups in this 3-part report (part 2 and part 3 here) - 37% say AI lowered CAC and 72% improved upsell/cross-sell.  Those Startups dedicating 50%+ of their GTM tech stack to AI hit meaningful scale: higher ARR brackets, and outsized efficiency.  Some major barriers remain for many of us: 23% cite high cost, 17% cite a lack of AI expertise.


3. GROWTH LEVERS: Fast follow from number 2 above, a lot of startups say AI growth is now cheap. Kyle Poyar's Growth Unhinged shows how top PLG teams spend less on new channels, yet sell more via AI-driven content, intent-outbound, and search. If you're not treating AI as your growth stack, you're probably falling behind.


4. MOATS: A16z revisits their 2020 classic: margins don't make great companies, moats do. In this new AI-infused era, defensibility still comes from the old pillars - network effects, brand, scale, switching costs, proprietary tech - and now, momentum reeeeaaallly matters too. High margins alone can be a red flag that your product isn't using AI.


5. WRITING: Here is a bit that flips on the AI panic: it's not writers getting replaced, it's readers. It's kinda like SEO - writers start optimizing for LLMs as their true audience, trading human readers for influence and speculative digital immortality....let the enshitifcation or dead internet theory flow!


6. HIRING: Hey founders - here is a pretty accurate techism: You will never be able to hire anyone better than you. Y'all set the talent ceiling, so curate your team like your product. Recruiting is the game.


7. NAME: Who knew that picking a company name could be complicated? Picking a company name looks simple - but most of us skip the hard parts. According to First Round Review's playbook, you should derive names from your positioning statement, brainstorm hundreds of options, test how your audience pronounces and remembers them, then vet for domain, trademark, and growth-fit.


8. AI WARS v2: I've talked about the Public Cloud AI wars underway in the past, and the Clouded Judgement blog has picked up on it too - Model scores don't rule anymore. With Gemini 3.0 that Google launched last week, Google isn't just chasing benchmark wins, it's bundling AI into Android, Chrome, Workspace, and more (like doubling AI capacity every 6 months!). Anyway, Google AI is all over my shit this week - here is how to turn it all off! The fight now is distribution, tools, and ecosystem -where we all touch AI every day.


9. VALUATIONS: There is a lot in here. The latest from Data Driven VC shows that top-tier private tech companies are losing ~20% of their value month over month; AI & crypto are still dominating; and EV/NTM revenue multiples have compressed significantly across sectors.


10. CASE STUDY: Check this deep-dive on Microsoft's AI strategy - which traces the "big pause" in datacenter build-out, how the Azure-OpenAI tie-up evolved, and why the real battleground now is tokens and infrastructure economics (see #8 above!).


POD OF THE WEEK: Lenny's Newsletter breaks down the clearest coldstart guide to building with AI: how to scope an AI product, when not to use models, evaluation traps, and what real teams ship first. If you're building anything AI-adjacent in 2026, use this to get your brain started.