Chapter 10 · Guardrails
Kill list, dos and don’ts
You can leave this chapter when you have crossed off every item you have done in the last 30 days and written the replacement.
Most stalled accounts are not lazy. They are busy at the wrong work. If an activity does not raise Gravity, it is a hobby. Hobbies are allowed. They are not the course.
Do
- Treat every reply as a mini-post
- Stay in one niche 90 days
- Publish a weekly number, even ugly
- Answer useful comments in 30 min
- Log profile visits, not just follows
- Refresh ten target names every Sunday
- Lock photo, bio job, and pin
- Talk to same-size accounts first
Don’t
- Bots, pods, follow/unfollow, engagement groups
- Paste a chatbot raw
- Feature changelogs as content
- Worship the 2023 13.5× table
- Train the graph on connect threads
- Hunt 200k accounts with “great point”
- Swap persona at 780 because you are bored
- Buy followers
| Looks like work | Do this instead |
|---|---|
| Redesigning the banner | Five-second bio test with a human |
| Planning 30 content pillars | Pick one job for 90 days |
| Reading 12 growth threads | Do the Chapter 4 drill |
| Building a scheduling stack | A calendar event named A/C block |
| Warm-up tools for a new alt | Rewrite the old account |
Myths, corrected
| Common claim | What holds up |
|---|---|
| “Never reply to big accounts.” | Half true. Use them for borrowed reach, in the first 15 minutes, with a real add. Peers stay the main effort. |
| “50 to 100 replies a day.” | Fine for a short sprint, not a lifestyle. Empty replies get ignored. Stay at 15 to 25 that pass the standalone test. |
| “Follow-for-follow is trash.” | Blind follow-for-follow is. Following back people you have traded real replies with is how most early growth actually happens. |
| “Video is the 2026 unlock.” | It matters more from 1k to 10k. A strong written explainer beats a mediocre video at this size. |
| “Just be consistent.” | Only with a kill list. Consistency on the wrong format keeps you stuck. |
| “Someone hit 200 in two days.” | Usually an existing audience or hours of daily replying. Do not use it to forecast yourself. |
| “Posting for months and nobody sees it means the algorithm hates me.” | Each post is scored on its own. Usually it is a reply, profile and friend-graph problem, not luck. |
| “A fake persona is a shortcut.” | Hallam tried one and got the follower spike, then said substance makes it easier. A persona with nothing behind it has nothing to keep people. |
| “Old 2022 to 2024 advice still works.” | Weights and formats have changed. Check any tip against your own last 14 days of data. |
| “Demos with long intros and small UI are fine.” | Slow opens, tiny visuals, clutter and long copy get scrolled past. Lead fast, show big, keep it readable without sound. |
Kindness is a filter, not a brand. Quote-dunks buy reach and spend Gravity. If the only people who reply are enemies, you will not have a Column C.
Drill · 10 minutes
List three things you did last week that belong on the don’t list. Write the replacement. Put it on tomorrow’s card.