KLIK · CREATIVE PRODUCTION · 2026

Your performance team needs 50 ad concepts a week. Your creative team ships 8. We close the gap two ways — we ship the creative, or we ship you the system that ships it.

Creative
production,
by the [klik].

60/wk
CONCEPTS SHIPPED
3 DAYS
BRIEF → LIVE AD
−60%
PRODUCTION COST
2 PATHS
DFY · OR · DIY

We used to buy media.
Now we make
the creative.

For nine years we ran paid acquisition. Media buying was the wedge — creative was always the bottleneck. We watched the same scene a hundred times: the buyer begs the studio for variants, the studio ships eight, the algorithm eats them by Tuesday.

So we built the pipeline ourselves. Tag every ad. Score every brief. Ten copy agents iterate on the script until it hits 90+. Image and video prompts come out the other end. Eighty concepts a week, ranked, every Monday.

Now we sell it two ways. We ship the creative for brands that want a production studio on retainer. Or we ship you the system — same flows, same prompts, full source — for in-house teams that want to own the pipeline themselves.

“You don't need more designers. You need a faster loop.”

— DALAI LAMA · DEC 2025

Pick a path.

Same pipeline behind both. The only question is who clicks the button — us, or your team. Most clients start with WE SHIP for a quarter, then graduate to YOU SHIP once they've seen the system work.

OROR
▶ PATH A · DONE-FOR-YOU

We ship.

Creative production on retainer. We run the pipeline, hand you 80 ranked concepts a week, ready to upload. Static, motion, scripts, hooks. You approve, your buyer launches.

FROM
$6K
/ MO · 3MO MIN
  • 25 concepts / week · scored & ranked
  • Statics, scripts, motion stills, hooks
  • Weekly brief drop · Monday 9am
  • Tagged + sorted in your Notion / Drive
  • Slack channel · 48hr turns
  • Monthly performance read-out
★ BEST FOR

DTC brands & agencies $75K+/mo paid creative spend who want output, not a system to learn.

HIRE THE SHOP ▶
▶ PATH B · BUY THE SYSTEM

You ship.

The same pipeline, transferred. Source prompts, scoring logic, brief tree, orchestration code — installed in your stack, owned outright. We embed for six weeks. After that you run it.

FLAT
$9K
ONE-TIME · 6 WK INSTALL
  • Full system source · prompts + code
  • 6-wk embed with your team
  • Loom library · 30+ training videos
  • SOPs · brief, score, ship pipeline
  • Runs on your own API keys
  • Optional $500/mo tune-up after
★ BEST FOR

In-house teams & agencies who already have producers and want to own the IP forever.

BUY THE SYSTEM ▶

Six modules.
One flow.

The same six modules sit behind both paths. On Path A we run them for you. On Path B we install them in your stack and walk away.

01 · TAGGING

Ad tagger

Every ad you and your competitors run, tagged by persona, angle, hook, format, and CTA. Reads Meta Ad Library + your own creative folder. The intel layer everything else feeds off.

02 · BRIEF TREE

Hypothesis
generator

Turns the tagged library into a ranked brief tree — P1 through P4 — with explicit hypotheses, target persona, and the angle each one is testing. The thing your strategist used to spend a Friday on.

03 · COPY ENGINE

Script + hook
scorer

Ten copy agents iterate every script until it scores 90+ against the angle's hit-rate prior. Hooks scored separately. Anything below threshold doesn't leave the pipeline.

04 · IMAGE PROMPTS

Visual
prompter

Each approved script gets matched image prompts — Nano Banana Pro 2, Midjourney, or your in-house gen model. Brand-consistent, persona-targeted, ad-format aware (1:1, 4:5, 9:16).

05 · VIDEO CUTS

Motion
pipeline

Storyboard prompts for Seedance / Runway / Higgsfield. Cut on a 4-beat structure (hook · problem · proof · CTA). Auto-captioned with Whisper, ready for upload.

06 · DELIVERY

Drop loop

Monday 9am drop in your Notion + cloud folder. Tagged, sorted, scored, with the original hypothesis attached. Your buyer pulls what they need; we log what shipped.

Why not
just hire?

Fair question. Here's how Klik compares to the three options you'd weigh — a creative agency, hiring in-house, or stitching together your own AI stack.

CRITERIA
KLIK
CREATIVE AGENCY
IN-HOUSE TEAM
DIY AI STACK
CONCEPTS / WEEK
2+ · scored & ranked
8–12 · eyeballed
10–20 · varies
Unbounded · unranked
RAMP TIME
2 wks → first drop
3–6 months
4 mo hiring + ramp
Months of trial-and-error
MONTHLY COST
$12K+ DFY / $48K one-time
$40–80K retainer
$25K+ salaries & stack
$2–5K tools + opportunity cost
SYSTEM OWNERSHIP
Yours on Path B
×Agency owns the IP
Yours · eventually
Yours · if you finish it
QUALITY CONTROL
90+ scoring · brand-trained
Senior CD review
Depends on the hire
×You QA every output
EXIT
Quarter-by-quarter
12-mo lock-in typical
Hard to fire
Sunk time · no resale

Four steps,
two endings.

Steps 1 and 2 are the same on both paths — we diagnose your library and install the pipeline. Steps 3 and 4 fork based on which path you picked.

STEP 03+ FORKS ▶ PATH A · WE SHIP · OR · PATH B · YOU SHIP
WK 01–02 · BOTH

Diagnose

Audit your last 90 days of creative + spend. Tag every ad. Map the hypothesis tree against the gap.

WK 03–06 · BOTH

Install

Build the six modules in your stack — your Notion, your folder, your accounts. Same pipeline, both paths.

WK 07+ · PATH A

We ship
weekly

80 concepts a week, every Monday. We tune prompts, kill what doesn't work, scale what does. Indefinite.

WK 07–12 · PATH B

You take over

Six weeks of embed. Your team learns the loop. Loom library, SOPs, prompt source transferred. We leave.

Common asks.

What's the actual difference between Path A and Path B?+
Who runs the pipeline. On Path A we run it from our shop and drop 80 finished concepts in your folder every Monday — you upload, we iterate. On Path B we install the same pipeline in your stack, embed for six weeks training your team, then hand over the keys. Same modules, same scoring logic, same output quality. The choice is whether you want output or ownership.
Can we start on Path A and switch to Path B later?+
Yes — most clients do. A quarter on Path A gives you time to see if the system actually performs in your category before you commit to bringing it in-house. If you upgrade to Path B within 12 months, we credit 50% of your Path A spend toward the install fee.
Do we own the prompts and source code on Path B?+
Yes, entirely. Source prompts, scoring logic, orchestration code, and SOPs are all transferred at handoff. The system runs in your accounts — your OpenAI, your Anthropic, your image gen — under your enterprise terms. We don't gate access through SaaS we own. After the embed ends, you're free to fork it, sell it inside your agency, or never speak to us again.
How much human creative judgment is left in the loop?+
The expensive parts. Strategy, taste, and the final approve/reject call stay human. The pipeline replaces the layers between brief and export — the copy iteration, the variant generation, the format adaptation. On Path A our creative director approves before drop. On Path B yours does. The model handles the throughput; the human handles the bar.
Is this a fit for a 5-person agency or only for brands?+
Both — agencies actually get the sharpest economics. If you service 5+ paid-media clients, one installed Path B pipeline runs all of them and your margin on creative production roughly triples. For a single brand the question is whether you're spending enough on paid creative ($75K+/mo) for the throughput to matter.
What does our data leave the building for?+
Only what you approve. Tagging runs against your own creative library + public Meta Ad Library data. Generation calls the model providers you already use under your own enterprise terms. We don't proxy anything through Klik accounts — your data, your keys, your audit trail.

Try one
module first.

Easier to play than to read about. Drop your email and we'll send the hook-generator module — one of the six in the pipeline, runs in Claude Code, full source, no upsell. → full cartridge spec

★ FREE · v1.0

HOOK-1
brief → 12 hooks.

Drop one line of product context. Get twelve testable opening hooks across four angles — pain, curiosity, authority, outcome — scored and ranked against Meta hit-rate priors.

It's a Claude Code skill. One markdown file. Runs locally on your own API keys. Source is yours, no SaaS in the loop.

  • RUNTIMECLAUDE CODE / SKILL
  • FORMAT.MD · 1 FILE
  • OUTPUT12 HOOKS / RUN
  • PRICE$0
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.

ONE EMAIL, ONE LINK, DONE. NO LIST.

~/claude/skills/hook-1.md
---
name: hook-1
description: brief → 12 ranked hooks
module: 3 of 6 · klik pipeline
---

# Hook generator v1

Input: one line of product context.

Output: 12 hooks across 4 angles
  · pain      · curiosity
  · authority · outcome

Score: 1–10 vs. the angle's hit-rate
prior from 90d of Meta data.
Cut anything below 7.

$ claude run hook-1 \
    "kettle & fire bone broth"
→ 12 hooks scored, top 3 starred ★
▶ READY TO PICK A PATH

Let us ship.
Or own it.

If your team is spending $75K+ a month on paid creative and the bottleneck is throughput, drop a one-paragraph note about the brand. We reply within two business days with a fit-check and a path recommendation.