Daily GTM Feed

5 Interesting Learnings from Klaviyo at $1.2 Billion in ARR: 32% Growth, 110% NRR, and Somehow Only 4x Revenue

SaaS dead, dying or underpriced? Feels like a stock pickers market with attractive opportunities to me

• Klaviyo trading at 4-5x revenue despite 32% growth, 110% NRR, and profitability—potentially most mispriced public B2B company or signal of 'New Normal' for SaaS valuations
• NRR improved to 110% while scaling to $1.2B ARR by doubling $1M+ ARR customers and growing $50K+ customers 37% YoY—rare upmarket expansion success at scale
• International revenue grew 42% YoY and now represents 33%+ of business, breaking 'Shopify add-on' narrative with regional hubs in Dublin and Singapore

The slow decay of growth (and how to avoid it)

• Growth decay is a common pattern affecting successful PLG companies including Ramp, Notion, Airtable, Figma, Miro, and Canva
• There are documented examples of companies that successfully reversed growth deceleration
• Newsletter promises new data and real-world frameworks for addressing growth plateau

Mythos, Muse, and the Opportunity Cost of Compute

brilliant read

• Reasoning models (o1) fundamentally break Aggregation Theory by reintroducing marginal costs - compute scales with usage, unlike internet-era products
• Hyperscalers' business models were built on zero marginal cost assumption; AI inference costs challenge this foundation requiring new economic models
• The 2010s internet era may be viewed as anomalous 'naive time' - technology returning to capital-intensive, high-marginal-cost paradigm of pre-internet era

AI Weekly Issue #481: Musk wants Altman fired, Anthropic passes OpenAI, Meta goes closed

• Anthropic overtook OpenAI in revenue ($30B vs $24B run rate) driven by enterprise customers, doubling million-dollar accounts in under two months
• Meta abandoned open-source AI strategy with first proprietary model under Superintelligence Labs, reversing Llama approach
• AI legal/regulatory activity intensifying: Musk-Altman litigation escalating, Hollywood writers secured four-year AI protections

OpenAI Forecasts Advertising to Hit $102 billion by 2030

• OpenAI projects advertising revenue to reach $102B by 2030, becoming its largest revenue driver
• Near-term forecasts show aggressive growth: $2.4B in 2024 to $11B in 2025 (4x increase)
• Represents strategic shift from subscription-first model to ad-supported monetization for AI platforms

How Databricks Sells to Dozens of Industries Without Building a Single Vertical Product

• Databricks uses 'imperatives' framework instead of traditional personas/ICP to sell horizontal platform across dozens of verticals without building vertical-specific products
• Imperatives sit at intersection of three elements: customer priorities (OKRs/accountability), industry trends (market movements), and your product capabilities (differentiated value)
• Traditional personas focus on buyer characteristics and ICP focuses on account fit, but neither forces you to speak the customer's strategic language or connect to their CEO's priorities

I have been coding for 11 years and I caught myself completely unable to debug a problem without AI assistance last month. That scared me more than anything I have seen in this industry.

Problem solving etc degrading is an interesting phenomenon. See it in education too😃, how do you build the competency in first place. Wonder what need is to intentionally not use Ai for tasks periodically?

• Experienced engineer (11 years) discovered degraded debugging ability after relying on AI tools - took longer to solve problem manually than would have 3 years ago pre-AI
• The 'internal monologue' that generates hypotheses under uncertainty atrophies with AI dependency - not just knowledge loss but fundamental problem-solving skill degradation
• GPS analogy: Using AI tools doesn't just provide answers, it prevents the formation of mental models that enable independent problem-solving when tools aren't available

I gave Claude Code our entire codebase. Our customers noticed. | Al Chen (Galileo)

So risky, but the grounding and ROI makes sense

• Non-technical field engineer eliminated engineering support bottleneck by giving Claude Code access to entire 15-repo codebase, creating self-service technical answers
• Code as source of truth beats documentation - current codebase provides more accurate answers than static docs, especially for complex multi-repo architectures
• Customer quirks system creates hyper-personalization at scale - combining repo context with Confluence, Slack, and customer-specific deployment patterns turns single questions into reusable knowledge

I built an AI job search system with Claude Code that scored 740+ offers and landed me a job. Just open sourced it.

• Individual contributor built sophisticated AI job search system using Claude Code that evaluated 740+ listings and resulted in Head of Applied AI role - demonstrates practical AI coding tool capabilities beyond simple automation
• System emphasizes quality over quantity with 10-dimension fit scoring to prevent spray-and-pray applications - contrarian approach to typical job search automation that prioritizes volume
• Open-sourced complete system (MIT license) with 14 skill modes including resume tailoring, company scanning, interview prep, and ATS optimization - shows emerging pattern of professionals building and sharing custom AI workflow tools

People anxious about deviating from what AI tells them to do?

I mean, true or not, real or not this is an interesting topic (more so than the actual link)

• AI over-reliance creating anxiety about deviating from AI recommendations even when contradicted by authoritative sources (manufacturer instructions)
• Emerging pattern of AI-generated authority superseding domain expertise and primary documentation in user psychology
• Critical gap in AI literacy: users not evaluating AI outputs against context-specific authoritative sources or applying critical judgment

Anyone else feel like AI security is being figured out in production right now?

• AI security is being figured out in production with enterprises running 300+ unsanctioned AI apps and most lacking dedicated AI security teams
• Attack patterns mirror early-stage tech adoption: prompt injection, over-permissioned agents, and shadow IT rather than sophisticated exploits
• Traditional security knowledge transfers incompletely - prompt injection ≠ SQL injection, agent permissions ≠ API auth - creating expertise gap despite emerging frameworks (OWASP, MITRE ATLAS, NIST)

some revops teams have stopped doing revops

The SaaS is dead, the SaaS is going to rebound argument well encapsulated in a revops reddit thread

• RevOps teams are misallocating resources by building AI-powered replacements for cheap SaaS tools, ignoring total cost of ownership including AI API costs, engineering time, and quality degradation
• The 'uncapped Claude budget' approach represents a new form of technical debt where teams optimize for visible SaaS line items while creating hidden costs in maintenance, bugs, and user friction
• This represents a broader pattern of AI tool misuse: teams are treating coding assistants as cost-saving measures rather than productivity multipliers, leading to false economies that burn more value than they save

An AI state of the union: We’ve passed the inflection point, dark factories are coming, and automation timelines | Simon Willison

absolutely brilliant and insightful interview

• November 2025 marked inflection point where AI coding agents crossed from 'mostly works' to 'actually works' - specific timeline for capability shift
• Three agentic engineering patterns for daily use: red/green TDD, templates, and hoarding - actionable framework for practitioners
• Dark factory pattern emerging: AI does its own QA without human code review - next evolution beyond current copilot paradigm

Brand optimization: What it is and why your AI visibility depends on it

Wonder how true this is that consistency matters for AEO and rankings in that way. How does that play wiht founders and creators?

• HubSpot is positioning 'brand optimization' as necessary for AI visibility (AEO - Answer Engine Optimization)
• Article frames traditional brand consistency work as critical for appearing in AI-generated responses
• Introduces concept of optimizing brand for AI discovery alongside human audiences - emerging marketing concern

The events and community playbook

• Personio Benelux transformed their Amsterdam office into a community hub over 3.5 years, hosting 11 unique event formats including HR breakfasts, executive dinners, and their own annual conference (HUG) with 1,250+ attendees across 7 offices
• The 'Stam Cafe' philosophy: open doors to anyone sharing company values, treating office space as community infrastructure rather than marketing cost center
• Multi-format approach included live LinkedIn talk shows, TED-style masterclasses (People Talks), and partner-hosted events (Pavilion, SaaStack, D2 collective, WeLoveSaaS, HR Tech Europe), creating ecosystem flywheel effect

How to turn Claude Code into your personal life operating system | Hilary Gridley

• Anti-system philosophy: Claude Code learns preferences through observation rather than upfront configuration, prioritizing simplicity over complex setup
• Yappers API approach: Verbal narration of work eliminates need for technical integrations—AI learns context by listening rather than connecting to apps
• 10x impact framework for task automation: Systematic method for deciding which tasks warrant AI automation versus human effort investment

[AINews] The Last 4 Jobs in Tech

• Org charts are fundamentally changing with AI - from AI Engineer (2023) to Tiny Teams (2025) to new role archetypes
• Yoni Rechtman's 99D framework proposes 4 core post-AI job categories for white collar tech work
• Meta has officially adopted AI Engineer as an org structure, validating the 2023 prediction and showing enterprise adoption of new AI-native roles

Trust is the New Currency in B2B Buying: SurveyMonkey, Reddit

These are high % stats showing what we implicitly already know

• Peer validation (73% trust) now dramatically outweighs traditional vendor marketing (55% trust vendor sites, 39% trust AI chatbots, 36% trust social media) in early-stage B2B buying
• 83% of B2B buyers complete self-directed research before sales engagement, with high-stakes categories (software, professional services, HR) taking several weeks to months in extended evaluation
• Search engines serve as navigation layer, not destination—buyers use search to identify options then validate through peer communities like Reddit (121M daily users, 19% YoY growth), creating imperative for authentic community presence

Claude Channels

The move from user initiated to automated workflows is one of the main transitions with current agentic capabilities IMO

• Claude Channels (launched March 20, 2026) enables event-driven AI automation via MCP protocol, shifting from pull-based (user-initiated) to push-based (event-triggered) workflows
• Practical use case: CI/CD failures can trigger autonomous investigation, fix deployment, and resolution without human intervention - reducing 12-hour incident windows to near-zero
• Technical implementation uses MCP servers connecting Claude Code to messaging platforms (Telegram/Discord at launch), with Bun runtime for 4x faster cold-start performance vs Node

AWS Accelerates Internal AI Agents Following Staff Cuts

If you think white collar job displacement is a joke, or a distant future concern, this is just one more sign it is most definitely NOT. It's here.

• AWS is deploying AI agents to handle technical sales support functions previously performed by thousands of specialists
• The AI automation directly correlates with recent layoffs of hundreds in sales, business development, and technical specialist roles
• Major cloud provider is using its own AI capabilities to reduce headcount in customer-facing technical roles, signaling broader industry trend

AI News Weekly - 100 years from now : The Case for Artificial Stupidity - Mar 23rd 2026

now this is truly thought-provking

Why Hasn’t AI Made Work Easier?

• Large-scale study (164K workers, 180-day tracking) shows AI adoption doubled time spent on email/messaging/chat and increased business software use by 94%, but reduced focused work time by 9%
• This represents a 'productivity paradox'—AI accelerates shallow, context-switching work while cannibalizing the deep work that drives actual value creation
• Pattern repeats historical technology adoption cycles (email, mobile, video-conferencing) where efficiency tools paradoxically increased busyness without proportional output gains

Cursor admits its new coding model was built on top of Moonshot AI’s Kimi

• Cursor's new coding model is built on Chinese AI company Moonshot AI's Kimi foundation model
• This represents a significant supply chain transparency issue in a widely-adopted developer tool
• Geopolitical tensions around Chinese AI models create regulatory and compliance risk for enterprises using Cursor

Confronting the CEO of the AI company that impersonated me

• Grammarly/Superhuman shipped 'Expert Review' feature that cloned real journalists and experts as AI advisors without permission, triggering class action lawsuit
• Company response evolved from email opt-out to killing feature entirely after backlash - demonstrates reactive vs proactive approach to AI ethics
• Case illustrates emerging regulatory/legal risk category: AI companies using real people's professional identities and expertise as training data or product features without consent

Where Were YOU When the Revenue Multiples Died?

The GTM Engineer Pulse | #20

this is an interesting acquisition and consolidation, dont bet against apollo still

• GTM platform consolidation accelerating: Apollo acquiring Pocus signals aggressive upmarket push and category consolidation, with Apollo positioning as 'AI-native operating system for GTM teams from SMB to enterprise'
• AI agent infrastructure going 24/7: Anthropic's scheduled tasks for Claude Code enables always-on AI agents, backed by $100M ecosystem investment — fundamental shift from on-demand to autonomous GTM operations
• Terminal-native GTM emerging: Relevance AI's Programmatic GTM from terminal represents new paradigm where GTM operations become code-first, developer-driven workflows rather than UI-based point solutions

The bottleneck flipped: AI made execution fast and exposed everything around it that isn't

• AI compressed execution speed (weeks to hours for prototyping) but exposed coordination/decision-making as the new bottleneck - approval chains, planning cycles, and leadership velocity didn't accelerate
• 55% of CEOs who cut headcount citing AI already regret it; 42% of companies abandoned AI initiatives in 2025 (up from 17% prior year) - suggesting premature optimization and misdiagnosis of productivity gains
• Monday.com's counter-strategy: automated 100 SDRs but redeployed instead of cutting, recognizing 'every time we eliminate one bottleneck, a new one emerges' - treating AI as bottleneck-shifter not headcount-reducer

What Wins When Anyone Can Build Anything with Brett Queener, Partner at Bonfire Ventures

would echo field marketing is really the lifeblood now

• Traditional GTM competitive advantages (explanation, education, sales motion sophistication) are being commoditized by AI - 'right to win' now resets every 30 days
• Events creating 75%+ of pipeline for early-stage companies as face-to-face becomes critical when buyers bet careers on products that didn't exist 6 months ago
• Vertical software more defensible than horizontal platforms in AI era; Salesforce/HubSpot ecosystems risk becoming 'boat anchors' as agentic software collapses traditional sales motions from SDR to CSM

AI Native or Death. Choose.

Super useful and interesting — core AI-native thesis directly relevant to STEEPWORKS positioning. Victor flagged manually.

• AI-native companies show dramatically higher revenue per employee than traditional SaaS
• The AI-native label requires specific product characteristics, not just AI adoption
• Financial metrics reveal a clear performance gap between genuine AI-native ventures and AI-washed incumbents

Built an autonomous system where 5 AI models argue about geopolitical crisis outcomes: Here's what I learned about model behavior

• Multi-model consensus systems reveal significant disagreement (25+ points) between leading AI models on identical scenarios, with Grok showing bias toward OSINT signals
• Models anchor to their own previous outputs when shown historical context, requiring 'blind' operation to maintain independent reasoning
• Grounding/RAG prevents source hallucination but not content hallucination—models can fabricate specific data while correctly citing authoritative sources