Rosewater Creative Playbook
How to build October's first batch.
- Volume comes from variation
- Base concepts in a cluster
- The 3 Hooks theory
- How to build a hook
- Putting it together
Volume comes from variation, not from more ideas
A monthly creative number is not a list of that many separate ideas. It is a smaller number of base concepts, each built out.
Teams stall because they try to invent the whole number. Thirty six good ideas is a hard month. Three hundred and fifty good ideas is not a month, it is a career. The teams who hit their number are not more inventive than everyone else, they have built a system of variation on top of a normal number of ideas.
The three multipliers
Each one turns one idea into several adsCopy and imagery
Hold the format, change the headline promise and the hero image. One layout, five different promises to five different people.
Three hooks
Change the first ten seconds of a finished video and keep the rest of the cut. One video becomes three ads.
Format and placement
The same concept as a static, a carousel and a vertical cut. The idea is the same, the container is not.
Not hitting the number in full is fine. The number is a direction, not a test you pass or fail. A month that lands at 120 of a planned 180 has still put 120 new ads into the auction, which is more than enough to learn from. The goal is to get closer each month, and the way you get closer is by building the system once rather than sprinting every month.
One concept, five assets, one layout
Hold the format constant and change the big things: the copywriting headline and the imagery.
Same layout, same type hierarchy, same wordmark position, same offer lockup, same 4:5 crop. Five different promises to five different people, each paired with imagery that shows that promise. Click any frame to open it larger.
- Layout and safe areas
- Type hierarchy and weights
- Logo or wordmark placement
- Offer lockup and its position
- Aspect ratio
These are the parts that cost time and teach you nothing. Lock them once.
- The headline promise
- The hero image
- The proof element
- The colour field behind the subject
These are the parts the auction actually reads. Change them properly.
Take one layout that has already worked. Write five headlines that each make a different promise to a different person. Pair each headline with imagery that shows that promise rather than decorating it. Ship all five.
A variation is only a variation if the promise changes. Five ways of saying the same sentence is one ad, not five.
Change the first ten seconds, keep the rest of the cut
One finished video becomes three ads. Body, proof, offer and call to action all stay exactly where they are.
The expensive part of a video is everything after the hook. The shoot, the edit, the grade, the sound and the approvals all live in the body. Once that is paid for, two more ads cost you two more openings. This is the cheapest volume available to any team.
Four rules that make it work
Decide these before the shootWrite the three openings before the shoot
Not after. If the openings are written first, they get shot properly on the day instead of being cut out of leftover coverage.
Keep the cut point clean
The body has to start in a place that any of the three openings can hand off to. No dialogue running across the join, no music phrase that only resolves from one opening.
Make the three genuinely different
Three approaches, not three phrasings of one line. A walk-in, a product reveal and an on-screen promise are three hooks. Three versions of the same voiceover are one.
Test them as three ads
Three separate ads, so each opening gets its own delivery and its own read. Buried inside one ad as three versions, you learn nothing about which opening earned the result.
Four devices, and what each one looks like on screen
A hook is a mechanical decision, not a flash of inspiration. Pick a device and build it.
Movement
Open on motion so the frame changes inside the first second. Walking toward the camera is the reliable version: the subject starts far and arrives, so the frame is still changing at the moment the viewer decides whether to stay.
Visual hook
Something happens on screen rather than something being described. The product doing the thing beats the product sitting still. A state change, a reveal, a before and after inside two seconds.
Copywriting on screen
The promise is burned into the frame during the hook so it reads with the sound off. Most of the feed is muted, so a hook that only works with audio is a hook that works for a minority of impressions.
Editing
A visible cut or pattern break inside the first few seconds. The edit itself is the hook: the rhythm change is what tells the viewer something is happening here.
What usually goes wrong
Easy to spot in a kit reviewPutting it together for October's first batch
Six decisions, in order, from the cluster list you already have.
Start from your cluster list
The clusters are already set. This is not the moment to reopen them.
Pick the base concept per cluster
One concept per cluster. One is enough, because the concept is about to become five or eight ads.
Decide the format
Static, carousel or video. Choose the one the concept needs, then lock the layout so it stops being a decision.
Write the five headline promises
Five promises to five different people. If two of them could swap headlines without anything changing, you have four.
Book the imagery
One hero image per promise, showing the promise rather than decorating it. Book all five in one session.
Work back from the kit date
Kits are approved five days before go-live, so the go-live date is what sets the brief date.
The go-live date sets the brief date, not the other way round. Kits are approved five days before they go live, which means a Monday 5 October launch needs the kit signed off on Wednesday 30 September and briefed the week before. Work backwards from the date you want the ads live and the rest of the calendar writes itself.
One cluster, one week, eight ads
Worked example · any brand| Date | Step | Output | Ads in hand |
|---|---|---|---|
| Mon 21 Sep | Pick the base concept and lock the format | One concept, one proven layout | 0 |
| Tue 22 Sep | Write the five headline promises | Five promises, five people | 0 |
| Wed 23 Sep | Book imagery and write the three video openings | Five hero shots, three openings | 0 |
| Thu 24 to Tue 29 Sep | Shoot and build | Five statics, one video body, three hook cuts | 8 |
| Wed 30 Sep | Kit approved, five days before go-live | Eight ads approved together | 8 |
| Mon 5 Oct | Live | Eight ads from one concept and one shoot | 8 |
Run this for each cluster and the monthly number stops being something you invent and starts being something you schedule.
Rosewater plans the batch, writes the briefs, schedules the kits and measures what happens once they are live. Your creative team and your production partners build the ads. This page is the shared language between those two halves, so a brief and a kit mean the same thing on both sides.