Skip to content
Drako Podcast

Episode 04 · Campaign automation design · Cross-regional

Agents in the Journey

Campaign automation design in the agentic era — design the triggers, not the tools, and measure the machine against CAC.

This episode is scheduled. Audio, transcript and video highlight publish with the episode.

Why this episode, why now

Campaign automation used to mean a drip sequence and a handful of if-this-then-that rules. In 2026 the lifecycle platforms a startup is likely to run — Klaviyo, Braze, Customer.io, HubSpot, ActiveCampaign — ship agents: message-level personalisation, journeys that rewrite themselves, decisioning inside the flow, and MCP server support so other agents can call them. The comparison literature now treats those features as table stakes on upper tiers rather than differentiators, and splits the market into two camps: incumbents that bolted agents onto existing journey engines, and AI-native tools that treat the agent as the orchestrator and the channels as tools it calls.

The economics have moved with the features. Companies deploying AI agents for lifecycle email, ad copy generation and SEO content production report CAC payback three to five months shorter than non-adopters, per ICONIQ data; product-led companies acquire customers for under US$500 in most segments while sales-led equivalents run into the thousands; and HubSpot's switch in April 2026 to outcome-based pricing for its Breeze agents — a dollar per qualified lead, fifty cents per resolved conversation — means automation now has a unit cost a founder can put straight into a CAC model. Scott Brinker's 2026 prediction adds the other half: the shift is from AI for marketers to AI for customers, with buyers arriving through agentic browsers that have already done the comparison.

That is the setting for the fourth stage of the Drakopoulos Ventures framework, campaign automation design: designing the automated campaigns and triggers that run acquisition at scale. The case on the firm's site — a SaaS platform scaling user acquisition through an automated funnel architecture, tracked against conversion and customer acquisition cost and reviewed quarterly — is the practical spine. The argument of the episode is simple: in an agentic stack the scarce skill is not configuring a platform, it is deciding what should trigger what, where a human stays in the loop, and which number the whole machine is accountable to.

Daniel hosts from the seam he has always worked — Marketing and Information Systems by training, agencies and seed-stage startups in London, Singapore and Sydney by career, and a firm since 2017 built on marketing and engineering running to one plan. Automation is where that alignment either exists or is exposed.