Playbook / Draft

The playbook, one step at a time.

Nine steps, each ending in a decision and a concrete artifact. We develop one step fully before starting the next. Right now only Step 0 has a draft.

Draft

Proposed, not adopted. Nothing on this page authorizes spending, a pilot, or publishing. Founder-time and cash caps are still unset.

01 / Map

The nine steps

Episodes normally cycle through Steps 1–7. Step 0 and Step 8 work at the system level. A failed check sends us back to the step that owns the problem.

0 Define the channel Proposed

Focus: audience, editorial promise, platform roles, and limits.

Produces: a content charter and a platform matrix (both drafted below).

Done when: the founder records the goal, audience, platform roles, time cap and cash cap, and the upload routes and mobile previews are checked.

1 Select and validate the idea Not started

Focus: turn a subject into a specific viewer question and run the cheapest useful demand check.

Produces: an idea card with the viewer question, audience signal, original angle, and one piece of evidence we can show.

Decision: GO, REWORK, or PARK.

2 Research and substantiate Not started

Focus: investigate the premise, try to disprove it, and separate verified facts from interpretation or simulation. Check rights early.

Produces: a source pack and claim ledger. Every consequential claim is supported, qualified, or cut.

3 Design the story and test the hard part Not started

Focus: hook, sequence and payoff, plus the cheapest test of whatever could make the piece fail.

Produces: a script or storyboard.

Decision: GO, REDESIGN, or STOP. The plain-diagram versus Claude Code + Lemo comparison belongs here.

4 Produce the episode Not started

Focus: build, edit, caption, and export. Log active work, rework, failed attempts, and render time separately.

Produces: a reviewable cut, editable sources, an asset and rights record, and a production log. It is ready for review, not for release.

5 Review release readiness Not started

Focus: watch the exact export a viewer will see. Check what the visuals imply, whether the central idea lands, and whether captions, audio, mobile legibility, rights and disclosure hold up.

Decision: READY, REWORK, or HOLD, tied to that exact version.

6 Package and publish Not started

Focus: final title, opening frame, description, source links, and platform disclosures. The package must promise only what the piece delivers.

Produces: a founder-approved release package and a record of what went live. Publishing stays a separate founder decision.

7 Measure and learn Not started

Focus: compare the result with the episode's goal and separate what we observed from what we suspect about the topic, packaging, opening or explanation.

Produces: a short postmortem with the evidence, its limits, and one change to test next.

8 Standardize and scale Not started

Focus: look across episodes for practices that held up. Decide what belongs in a template, a reusable component, or agent instructions.

Decision: REUSE, ADAPT, RETEST, or RETIRE. One successful render does not make a tool the default.

02 / Step 0

Content charter

Proposed

Primary goal

Find out whether original, mostly faceless technical explainers can help the right viewers understand a real software failure and what prevents it, using a workflow a solo founder can repeat.

YouTube Shorts is the primary discovery test. LinkedIn native video from the founder profile is a secondary test of whether the same explanation starts substantive conversations with people who face the problem.

Views, ad revenue and adopting Lemo are not the primary goal.

First candidate

"One order. Two webhook deliveries." It is a testable idea, not validated audience demand.

Audience (proposed): software builders and technical founders running SaaS or agent workflows, narrowed to one viewer segment per idea.

Promise: one concrete failure, one piece of evidence viewers can follow, and one practical takeaway. Simulations are labeled as simulations.

Constraints

03 / Step 0

Platform matrix

Proposed

Run YouTube Shorts as the discovery pilot and LinkedIn native posts from the founder profile as the business-response comparison. Do not launch six surfaces at once. Export sizes are our starting presets, not platform requirements.

Surface Role Starting preset What we measure Official reference
YouTube Shorts Pilot 1080×1920, 9:16, 35–50 seconds. One question, visible evidence in the opening, captions, one payoff. Engaged views and retention where available, not raw plays alone. Shorts classification
LinkedIn founder profile Comparison 1080×1350, 4:5, 30–60 seconds. Lead with the operational consequence; write the takeaway in the post. Substantive replies from target roles and qualified conversations, not generic reactions. Share videos
YouTube watch page Conditional 1920×1080, 16:9. A complete walkthrough; length follows the explanation. View duration and retention. Packaging tests only for eligible uploads. Aspect ratios
Instagram Reels Defer 1080×1920, 9:16, 30 fps. A self-contained visual story. Shares and saves where available. Reel size
TikTok Defer 1080×1920, 9:16. Start at the demonstration, not a logo. Watch time and useful comments. Do not budget for Creator Rewards without confirming eligibility. Creator Rewards
Facebook Reels Defer Start from the 9:16 master only if testing this audience. Audience settings, crop and account-specific metrics. Reels requirements
Shared release checks

Default to MP4 with H.264 video and AAC audio, 30 fps for the animated pilot, a clean narration track, and an SRT captions file. These are project presets, not universal limits.

Make separate compositions when framing changes; do not crop out evidence. Preview captions, labels and the cover on a phone with sound on and off. Record rights for visuals, fonts, voice and music per surface, and verify each platform's disclosure rules. Check account region, metrics availability and the upload route before any release decision.