The AI Growth Marketing Playbook: What's Actually Working in 2025

AI in growth: the productivity paradox

Everyone has access to the same AI tools. So why are some growth teams 3× more productive while others have added AI to their stack but aren't seeing results? The answer is almost always workflow design — not which tools they chose.

What high-performing growth teams are doing

Programmatic content at scale, with quality gates

The teams winning at AI-assisted content aren't publishing everything Claude generates. They're using AI to generate 10× the volume, then using evals (often Claude-as-judge) to filter to the 20% that meets their quality bar. The result is more output at consistent quality, not lower-quality output faster.

AI-assisted lifecycle automation

Lifecycle sequences (onboarding, re-engagement, expansion) that used to take weeks to design and write are now built in days. More importantly, they're personalised at a segment level that wasn't previously feasible: different sequences for different ICP segments, industries, use cases, and behavioural signals.

SEO at the keyword cluster level

Instead of optimising individual articles, high-performing teams are using AI to map entire keyword clusters, generate content briefs for every important query, and build internal linking structures that capture traffic across the full funnel. Claude is used for brief generation; human writers and AI in tandem for the content itself.

Instrumentation before automation

The teams getting the most from AI growth tools have excellent attribution and analytics. They know which content is converting, which sequences are driving expansion, and which channels have the best CAC. AI amplifies what's already working — it doesn't rescue what isn't.