The moat moved.
Marketing got autonomous.

The thesis

A founder used to need a marketing team. Then a stack of tools. Now they need a system that acts — one that takes a goal, plans the campaigns, ships the content and ads, runs the experiments, reads the data, and decides what to do next without waiting for a human at every step.

Vibe coders can ship code trivially. The moat has moved to whoever can make people care and pay. Launchpad solves code-to-prod. Orbit solves product-to-paying-customer — and does it across an entire portfolio at once.

The moat — this is the whole point

The defensibility isn’t any one module. It’s the compounding loop only Galax can build:

Native product state

Launchpad built the app, so Orbit knows every button, plan, funnel event, and feature. Competitors stop at ad → landing. Orbit goes ad → landing → signup → activation → upgrade → retention in one closed loop.

Cross-portfolio learning

Every experiment in every company writes to a shared playbook: hooks that converted, pricing curves that worked, lifecycle sequences that retained, channels that overperformed for ICP X. Company N+1 starts at the playbook’s current best, not from zero.

Founder voice IP

Fireflies transcripts plus writing samples become a voice model per founder. Generic LLM output dies on contact with audiences. Orbit’s output sounds like the person who built the company. Nobody else has this raw material flowing in.

Closed-loop execution

Orbit doesn’t just recommend — it writes to Stripe, Netlify, Meta, Google, lifecycle email. Decision and execution live in the same system, so cycle time is hours, not sprints.

More companies → more experiments → smarter playbook → better outcomes per company → more companies want in.

Most founders stitch these capabilities with Zapier duct tape. Yours speak natively.

How direction becomes autonomous execution

The founder doesn’t manage campaigns. They set direction:

• Objective. “$50k MRR by end of Q3.”
• Brand and voice. Voice samples, do/don’t list, taste rules.
• Budget and guardrails. Monthly spend ceiling, channels allowed, what needs approval.

From that, Orbit:

1. Plans. Decomposes the objective into a 90-day plan — channel mix, content cadence, experiment list, lifecycle program, budget allocation.
2. Executes. Ships content, runs ads, sends emails, tests landing pages, updates pricing — autonomously within guardrails.
3. Measures. Pulls every result into one funnel view: CAC, payback, ROAS, activation, retention.
4. Learns. Kills losers, doubles down on winners, writes new patterns to the portfolio playbook.
5. Briefs. Weekly: what I did, what worked, what I’m doing next, the one or two things I need you to decide.

The founder makes the high-leverage decisions. Orbit handles everything else. Target: ten minutes a day to run a company’s marketing.

The compounding asset

One structured store of:

• Every hook, headline, ad creative, subject line — paired with channel, ICP, and outcome.
• Every pricing test result.
• Every lifecycle sequence and its impact on activation and retention.
• Every voice sample per founder.
• Every “moment” that did or didn’t convert into traction.

After 3 companies it’s interesting. After 10 it’s a competitive weapon. After 30 it’s irreplaceable.

The relationship to Launchpad

Shared project identityis the bridge. From the founder’s perspective, they’re one system.

Orbit pulls from Launchpad: deployed URLs, feature list from the repo, Stripe plan state, Supabase funnel events, traffic and conversion data.

Orbit pushes back to Launchpad: landing page variants deployed through Netlify, pricing changes written to Stripe, lifecycle emails wired into auth flows.

Launchpad ships the rocket. Orbit keeps it moving.

What the founder does all day

Mondays: read the brief. Plan vs. actual. Top win. Top loss. What I changed. What I’m testing. What I need from you. Two approvals at most.

Tuesdays through Fridays: take calls, build the product, talk to customers. Open Orbit when an approval lands, when a reply is drafted, when a signal is hot enough to warrant a five-minute decision.

Saturdays and Sundays: the system is still running. The founder is not.

Ten minutes a day. Decisions, not chores.