Journal
Field notes from the automation stack.
Playbooks, audits, and opinions on performance marketing, AI attribution, and the automations we ship. No clickbait. No “10 tips.” Just what moves the number.
Google Ads for B2B SaaS in 2026: what still works
The playbook that compounds: intent keywords, tight negative lists, conversion events wired to closed-won, and a weekly roll-up the CFO can read in 30 seconds.
Meta Ads creative: the 90/10 rule that moves CPA
90% of Meta performance is creative, 10% is targeting — and most brands have those numbers inverted. A working framework for ad-creative velocity.
Brand strategy for performance marketers: the 30% that pays the other 70%
The CFO-defensible case for brand spend in a performance-first stack. How brand lift shows up in paid efficiency, organic share, and sales-cycle length.
AI attribution beats dashboards every time
Most teams stop at a ROAS number. We don't. Here's what pipeline-aware AI attribution actually looks like — and why it drives different decisions.
Which AI model should you pick for marketing in 2026?
A head-to-head on Claude 4, GPT-5, and Gemini 2 for the marketing workloads that matter: copywriting, creative brief generation, lead scoring, and email personalization.
ROI frameworks that survive a CFO review
Three-line P&L logic for marketing spend: incrementality-adjusted ROAS, payback period, and the 'kill when you can' rule. No attribution arguments.
AI agents in marketing ops: what actually works in 2026
Five agent patterns we've shipped to production — lead scoring, variant generation, reply detection, anomaly alerts, and weekly roll-ups — and what we tried that failed.
The lead-scoring automation we ship in week one
A concrete n8n + AI-agent playbook for scoring, enriching, and routing inbound leads — with the exact signals we weight and how we tune over time.
SEO in the AI-answer era: what still matters
Organic search changed. Zero-click and AI-answer boxes ate a chunk of traffic. Here's what still compounds in 2026 — and what to stop doing today.
Performance marketing benchmarks: B2B vs B2C in 2026
What good looks like across Meta, Google, and LinkedIn — CPL, CPA, ROAS, payback — segmented by B2B SaaS, B2B services, and DTC eCommerce.
LinkedIn outbound automation without getting flagged
LinkedIn's 2025 policy tightening killed most automation tools. What still works, what just got a 30-day ban hammer, and how we scale outreach safely for clients.
Channel ROAS consolidation: stop paying twice for the same click
The quiet ROAS leak in every multi-channel account: the same user, counted three times. Here's how we detect it and how much it typically saves.
Attribution: MTA vs MMM in the age of signal loss
iOS 17, third-party cookie deprecation, and platform walled gardens made MTA less reliable. MMM is back. Here's how to run both without going insane.
Meta ads on Android vs iOS: the signal gap that changes your bids
Meta's signal quality on Android is now materially higher than iOS for most B2C advertisers. What that means for bid strategy, creative allocation, and the weekly review.
Marketing org design for an AI-first team (2026)
The five roles you actually need in 2026. Why the 'growth marketer' and 'content marketer' titles are merging. How to structure for ship velocity without losing brand.