Split hook effort seventy, twenty, ten

1. The condition: this split needs results to work from

Sections 3 and 4 assume you have past content to look at. The split is for the second round onward.

If you are starting out, it is said to be even easier. Make a round with no idea what will happen. See what happened. Then follow the split from then on.

The promise attached: however bad your content is, some of it will land in the top 10% of that content.

• Look at the hooks on your top 10%. Make more like that.
• Look at your bottom 10%. Make less like that.
• Repeat, indefinitely.

That way you improve with no teacher at all; the numbers do the teaching. Most people have no idea anyway.

2. The first time through

A first pass that starts from other people's winners:

1. Look at the top-performing ads or content of other people.
2. Isolate their hooks.
3. Write down your favorite 50.
4. Use all of them in ads and content.

§7 sets the two side by side.

3. The split

The split is borrowed from a well-known innovation rule for dividing a company's resources. The claim that its originators proved mathematically that it gave them the best returns is made in passing; that proof is not established on this page.

Share	Bucket	What goes in it, for hooks
70%	Core	Hooks already proven, in your own content or in other people's: more of what works
20%	Emerging	Promising but less proven than the core: winner-adjacent ideas, such as a concept that works in another niche, modeled
10%	Big new ideas	Totally new things: new visual formats and hooks you think might work

70, 20 and 10 add to 100.

3.1 Seventy percent: the core

Put the largest chunk of effort into core activities: using hooks already proven in your own content or in other people's. The largest bucket goes on more of what works. That steadies all of your advertising and guarantees a baseline of performance — which matters when your staff want to be paid every month, not just when something takes off. For example, the best-performing posts on X get reused as hooks for short videos or captions.

The action step for the core: a hook spreadsheet.

1. Create a new spreadsheet.
2. Give every platform you run ads on its own tab.
3. Head the columns simply: name, hook, views and link.
4. Every time you go to make content, review your best hooks of all time.

3.2 Twenty percent: emerging

The next 20% goes to things that are promising but less proven than the core. For hooks, think of it as winner-adjacent: if something works in another niche, you may try to model the concept.

The action step for emerging: an outlier deck.

1. Find outliers. You can subscribe to software tools that show the outliers on each platform; search for them online. None is recommended by name. A few are in use, and each is a little different.
2. If you would rather not pay, search the platforms you are on by the topics you talk about, to find the highest-performing content.
3. Paste each outlier into a slide deck: a screenshot of it, and a few ideas for how you could repackage or reuse its formula.

3.3 Ten percent: big new ideas

The last 10% is for trying totally new things. The ideas here are risky, though they could lead to huge wins. This is where the craziest ideas get worked on, with no fear of failing. It keeps you ahead and keeps you thinking of new things to make: new visual formats, and hooks you think might work.

• If an experiment loses, document it so you do not repeat it.
• If it wins, it becomes a mainstay until something else beats it — it moves into the 70% bucket, more of what works.

Why split it this way. Most effort goes to what already works, some to new ideas that look promising, and a little to totally new things. That keeps money coming in now while keeping you ahead for later.

4. Every time after: the split as steps

The split as steps, for every round after the first:

1. Look at your top 10% of performers — the top 5.
2. Build 70% of the next round of ads on those five hooks.
3. Make 20% from permutations of those hooks, in visuals as well as words.
4. Make 10% completely different hooks, modeled the way you did the first time through (section 2).

The arithmetic: the top 10% is the top 5 when the round is 50 hooks, the number the first pass writes down.

These steps fill the 20% and the 10% differently from section 3. See section 7.

5. How many hooks to record, by kind of content

At the end of this you have large sets of hooks. The hook dictates the content, not the other way around, so you no longer need to worry about what content to make: you will have more than you can use. Turn the sets into recordings like this:

Content	What to record
Ads	10 or so hooks for every 1 piece of ad content
Short videos	Every winning tweet on every other platform is typically reused as a short video, finished by expanding on the hook
Long videos	3 to 4 hooks; the team picks in post-production which one they think will do better
Emails	Active split tests on every campaign sent, with the winners documented in a separate file

6. The ranking behind "top"

When the best hooks were collected, they were the top ones over the past year: content ranked by views, and ads and emails by performance. A second view ranks your own organic content by leads per piece rather than by views, tracking each piece with its own tagged link, because views do not say whether they came from buyers; the outliers by leads give a sense of which topics buyers respond to most. Any other measure of a top performer is not established on this page.

7. What this page does not decide for you

• Where the first round's hooks come from. §1 says to make a round with no idea what will happen, then keep what landed in your top 10%; it does not say where that round's hooks come from. §2 fills the first pass with 50 hooks taken from other people's top performers. Whether §2 is the way to make §1's round is not established on this page.
• What fills each share. §3 fills the 70% with hooks proven in your own content or in other people's, the 20% with winner-adjacent ideas (a concept seen working in another niche may be modeled), and the 10% with totally new ideas. §4 fills the 70% with your own top 5 hooks, the 20% with permutations of them, and the 10% with completely different hooks modeled like the first pass. Read the first way, the 20% may reach outside your niche. Read the second way, it stays inside your own winners. A third account, from one process example, splits a round of 50 hooks into 35 modeled directly on your top five winners, 10 small permutations of them, and 5 completely new pattern interrupts drawn from the wider market. This page does not settle which reading is right.
• How long a hook runs before it counts as a top or bottom performer. Not established on this page.

8. The checklist

Question	The answer
When the split applies	Once you have past content and results to look at
First time, starting out	Make a round with no idea what will happen, see what happened, then follow the split
First time, from other people's winners	Take other people's top hooks, write down your favorite 50, use them all
The split	70% core, 20% emerging, 10% big new ideas
Core tool	A spreadsheet: a tab per platform; name, hook, views, link; review it every time you make content
Emerging tool	A slide deck of outliers, each with a screenshot and ideas to repackage it
A new idea that wins	Moves into the 70%
A new idea that loses	Documented, so it is not repeated
Every round after	Top 10% (top 5): 70% reuse, 20% permutations, 10% completely different
Hooks per piece of ad content	10 or so
Hooks per long video	3 to 4
Emails	Split-test every campaign; document the winners
Top and bottom 10%	Make more like the top; make less like the bottom

9. What this page does not cover

What a hook is and how to write one is SOP 121 — Write a hook. Where winning hooks come from, and the weekly assembly of hooks, meat and calls to action, are SOP 119 — Assemble hook, meat and call to action. Hooks by level of awareness are SOP 118 — Write hooks for each level of awareness. The one-test-a-week cadence is SOP 41 — Run one test a week. Choosing between doing more, doing it better and doing something new for the next advertising effort is SOP 40, where new means a new placement, platform or method; the new hooks in this page's tenth are new creative inside channels you already run.

Budgets and the statistics of a split test are not covered on this page.