Work / Social Content
SocialContent
Finding the format first, then running it.
- DIRECTION
- PRODUCTION
- EDIT AND MOTION
- DISTRIBUTION
SEARCH
What is already working. Across the whole platform, not just my niche.
/ Overview
I ran five TikTok accounts to find out what actually works.
Not a content plan. A test bed. Search wide, ship small, and let the numbers pick the direction instead of picking it myself. Built and tested on my own products before it was ever offered as a service.
/ Project details
- Name:
- Social Content
- Role:
- Direction, production, edit
- Type:
- Content system
- Year:
- 2026
/ Challenge
Most content advice starts with a topic. Pick a niche, post consistently, be authentic. It skips the part that decides whether any of it works, which is format.
You cannot know the format in advance. You can guess, and most people guess once and then spend six months defending the guess. So the real question is how fast you can find out you were wrong, and how cheaply.
/ Provided services
Direction
A format worth repeating, found by testing rather than assumed at the start.
Production
A shoot setup that can run weekly, not a one-day circus.
Edit and motion
Templates and pacing built for a phone, so a series looks like a series.
Distribution
Multiple accounts running the same format in parallel, each with its own angle.
/ THE HOOK
The first three seconds decide the other fifty-seven.
Before
Here's my cheeseburger pasta recipe
After
This pasta tastes exactly like a cheeseburger.
// Lead with the claim. The recipe is the payoff, not the hook.
Before
5 tips for building muscle
After
You are eating enough. You are just eating it too late.
// Contradict something they already believe.
Before
How to track your calories
After
Nobody weighs their food forever. Here is what sticks.
// Name the objection before they think it.
Same video. Same effort. The difference is the line on screen when someone's thumb is already moving. This method produced over 4.5 million views in two months across five accounts.
/ Solution
I started with what was already working across the whole platform, not just my own corner of it, and tested concepts against each other in small batches. Nothing was precious. Anything that did not move got dropped inside a week.
One category pulled away: food and nutrition. So I stopped testing broadly and started testing inside it. Recipe content won, and once it won it stopped being a post and became a template, produced at volume across every account.
Search wide, test cheap, read honestly, narrow hard, then run the winner until it stops working.
/ What it produced
4.5M+
views in two months, after finding the format
5
accounts run in parallel
1
format, found by testing not guessing
/ What I shipped
A tested format
One winning concept, found by elimination, not by guessing.
Repeatable templates
Edit patterns and lockups so the next batch never starts from zero.
Five parallel accounts
The same format, different angles, all feeding the same read.
A hook process
Every piece written twice, shipped once.
