← Back to Blog

30 days, one Brain: a campaign-in-public retrospective

May 19, 2026 · 3 min read
30 days, one Brain: a campaign-in-public retrospective

We ran a 30-day campaign in full public view. Here's what it produced, what broke, and what the experiment proved about running content from a single Brain.

The short version

30 days, ~180 posts across 6 platforms plus blog, one operator, one Brain. This post is the retrospective: what worked, what broke, what the data said, and what changes in the next campaign.

The experiment

One month. One operator. One Brain. All content (drafts, scheduling, previews, analytics) inside a single Content OS.

No second tool. No manual writing. Just the 10-minute setup, the initial seed, and let the Brain run.

What the Brain produced

~180 posts shipped. Breakdown:

What worked

Three things compounded visibly: Knowledge Base (three uploaded docs made day-one output substantially sharper), Voice Drift Detection (caught three real drift events pre-publish), Closed-loop Performance Patterns (by week 3, hook quality measurably improved).

Pro tier was the right sizing for the workload. Founder would have been overkill.

What broke

Week 1: the Brain over-indexed on one pillar (product explainers) because the first high-scoring post was a product explainer. Performance Patterns corrected by week 3, but manual intervention sped it up by a week.

Two posts had voice drift we didn't catch pre-publish because the Voice Fingerprint was too loose at setup. Tightened the fingerprint, drift detection sharpened.

What the data said

Blog + LinkedIn combined drove most qualified signups. X drove traffic but thin conversion. TikTok drove reach but few signups. IG and Facebook were steady-state.

Next campaign: lean heavier into blog + LinkedIn. Experiment more aggressively with TikTok format.

One Brain. One surface. 30 days of compounding. The flex isn't the output. It's that the system runs without supervision.

Same offer, next campaign

7-day free trial. No credit card. Everything in this retrospective is reproducible inside Heist Pro at $49/mo.

The operation runs itself. Your turn to pull the Heist.

Steal your time back.

Heist is the AI Content OS that replaces your stack with one Brain that remembers, learns, and gets sharper with every post. A month of content in 10 minutes. Free 7-day trial, no credit card required.

Get Your Time Back →
FREQUENTLY ASKED QUESTIONS
What is the experiment?

One month. One operator. One Brain. All content (drafts, scheduling, previews, analytics) inside a single Content OS.

What does "What the Brain produced" cover in this post?

~180 posts shipped. Breakdown:

What does "What worked" cover in this post?

Three things compounded visibly: Knowledge Base (three uploaded docs made day-one output substantially sharper), Voice Drift Detection (caught three real drift events pre-publish), Closed-loop Performance Patterns (by week 3, hook quality measurably improved).

What does "What broke" cover in this post?

Week 1: the Brain over-indexed on one pillar (product explainers) because the first high-scoring post was a product explainer. Performance Patterns corrected by week 3, but manual intervention sped it up by a week.