A beauty brand approaches us with a familiar story. They spent £8,000 producing three hero videos for their new product launch. Professional production, professional talent, professional editing. They launched the videos with a £5,000 budget across Meta.
Three weeks later, the best-performing video had a 1.2% CTR and a 1.4x ROAS. They killed all three and started over.
The total cost of the experiment: £13,000 and three weeks of lost time during a launch window.
What went wrong was not the production quality. The videos were beautiful. What went wrong was the process. They produced the videos first, based on what they thought would work, and then discovered that their assumptions were wrong — after spending the money.
There is a better way. It is not complicated, but it requires inverting the order that most brands follow.
The three phases
The framework is simple: static, then video, then scale. Each phase has a specific purpose and a specific exit criterion.
Phase 1: Static testing. Produce 10-15 static variations testing different hooks, angles, offers, and product positioning. Launch them to your prospecting audience with minimal budget (£5-10 per static). Run for 48-72 hours. Identify the 2-3 hooks that generate the highest CTR. Total cost: £150-300. Total time: 3 days.
Phase 2: Video production. Brief and produce video creative around only the validated hooks from Phase 1. You are not guessing what will resonate — you already know. The static data told you. Produce 3-5 videos based on the winning angles. Total cost: £500-3,000 (depending on production quality). Total time: 1-2 weeks.
Phase 3: Scale. Launch the videos with confidence and meaningful budget. The hooks work (proven in Phase 1). The video execution brings them to life with demonstration, storytelling, and emotion. Scale the winners aggressively. Kill the underperformers. Total time: ongoing.
The key insight: Phase 2 is informed by Phase 1. The most expensive step (video production) only happens after the cheapest step (static testing) has validated the direction. This inverts the risk profile of creative production.
Phase 1: What to test and how to read it
The purpose of Phase 1 is not to find winning ads. It is to find winning messages.
Each static should test one variable — the hook. Keep everything else as consistent as possible: similar image quality, similar layout, similar call-to-action. The only thing that changes is the main message.
For a beauty brand launching a new hyaluronic acid serum, 10 static hooks might be:
- "72-hour hydration from one application" (benefit claim)
- "The serum dermatologists actually use" (authority)
- "Your skin loses moisture every hour. This stops it." (problem/solution)
- "3 types of hyaluronic acid. Here is why that matters." (education)
- "Why £32 outperforms £95 serums" (price anchoring)
- "12,000 women switched to this serum in 2025" (social proof)
- "The ingredient your moisturiser is missing" (curiosity gap)
- "Plumper skin in 7 days or your money back" (guarantee)
- "Everything your current serum does, without the irritation" (competitive)
- "This replaced 2 products in my routine" (simplification)
Launch all 10 at £7 each. Total spend: £70.
After 48-72 hours, the data reveals clear patterns. Maybe hooks 1, 5, and 8 all achieved CTRs above 3%, while the rest hovered around 1.5%. Now you know three things:
- Benefit claims with specific numbers work for this audience
- Price anchoring resonates (people care about value)
- Guarantees reduce friction
These three insights directly inform your video briefs. Hook 1 becomes a video that demonstrates the 72-hour hydration claim visually. Hook 5 becomes a comparison video showing your product vs. a premium alternative. Hook 8 becomes a UGC-style video leading with the money-back guarantee.
You are not guessing any more. You are producing.
Phase 2: Translating static winners to video
The translation from winning static to effective video is not a direct port. You cannot just read the headline on camera and call it a video. The static validated the hook — the video needs to bring that hook to life in a format that uses video's unique strengths.
Here is how each static hook type translates:
Benefit claims → demonstration videos. If "72-hour hydration" was the winning hook, the video should show the product being applied and then show skin hydration over time. Close-ups of skin texture. A time-lapse of the product absorbing. The video proves the claim that the static stated.
Price anchoring → comparison videos. If "£32 vs. £95" was the winning hook, the video should physically compare the two products. Show them side by side. Apply each one. Discuss the ingredients. Let the viewer see that the affordable option holds its own. The video delivers the evidence that the static promised.
Social proof → testimonial videos. If "12,000 women switched" was the winning hook, the video should feature real customers talking about why they switched. Multiple voices, authentic settings, genuine enthusiasm. The video humanises the statistic that the static presented.
Guarantees → founder or expert videos. If "money back" was the winning hook, the video should feature someone authoritative (the founder, a dermatologist, a formulator) explaining the guarantee directly to camera. Confidence, eye contact, specificity. The video builds trust around the promise that the static made.
Each video starts with the validated hook in the first 2-3 seconds — that is the part you already know works — and then spends the remaining 15-25 seconds delivering on it with video-specific storytelling.
Phase 3: Scaling with confidence
The brands that scale effectively on Meta share one trait: they know what works before they spend big.
Phase 3 is where the budget goes. Launch the videos from Phase 2 with real budget — not £7 per ad, but £50-200 per day depending on your overall spend. Give the algorithm enough data to optimise.
Within the first week, you will see which video executions perform best. Scale the top 1-2 videos aggressively (increase budget by 20-30% every 2-3 days while ROAS holds). Pause the underperformers.
Simultaneously, keep the winning statics from Phase 1 running in retargeting and catalogue campaigns. They have already proven their message — now they serve the mid-funnel while the videos handle cold acquisition.
The result is an account where every piece of creative, video and static, is data-informed. Nothing is live on a hunch.
Applying the framework to different scenarios
The static-video-scale framework works across different campaign types, with adjustments:
Product launch:
- Phase 1 (week 1): Test 15 hooks around the new product — different benefits, different positioning angles, different audience framings
- Phase 2 (week 2-3): Produce 5 videos around the top 3 hooks
- Phase 3 (week 3+): Launch at scale, time the peak spend with the launch announcement
Seasonal campaign:
- Phase 1 (4 weeks before): Test 10 seasonal offer variations — bundle pricing, gift with purchase, limited editions, discount structures
- Phase 2 (2 weeks before): Produce 3 videos around the winning offer
- Phase 3 (campaign window): Launch at scale with seasonal budget
Always-on acquisition:
- Phase 1 (ongoing): Run 5-10 new static tests every two weeks, continuously testing new hooks and angles
- Phase 2 (ongoing): Produce 2-4 new videos monthly based on the latest winning hooks
- Phase 3 (ongoing): Scale winners, retire fatigued creative, maintain a portfolio of 5-10 active video ads at all times
The cadence changes but the sequence never does. Static first, always.
Common mistakes
Skipping Phase 1 because "we already know what works." You do not. The data consistently surprises even experienced media buyers. A hook that sounds brilliant in a brief can fall flat in the feed, and a throwaway angle can be the biggest winner. Test everything.
Testing too few variations. Ten statics is the minimum for Phase 1. With fewer, you do not have enough variation to identify clear patterns. With 15-20, you increase the chances of finding an unexpected winner.
Spending too much on Phase 1. This is a screening round, not a performance campaign. £5-10 per static is sufficient. You need enough data to see CTR patterns, not enough data to calculate precise CPA.
Producing videos that ignore the static data. If the winning hook was "72-hour hydration" and the video opens with a 10-second brand logo animation before mentioning hydration, you have wasted the insight. The hook that won in static must be the first thing the viewer sees in the video.
Not running Phases 1 and 3 simultaneously. While your current videos are scaling in Phase 3, you should already be running Phase 1 tests for the next round. The framework is cyclical, not sequential. There is always a static test running, always a video being produced, and always a winner being scaled.
Inverting the risk
The traditional creative workflow for beauty brands is: have an idea, produce a video, launch it, hope it works. The cost of a bad idea is £1,000-10,000 and weeks of lost time.
The static-first workflow is: have ten ideas, test them as statics, identify the winners, produce videos of only those winners, launch with confidence. The cost of a bad idea is £7 and 48 hours.
Same creative ambition. Same production quality. Radically different risk profile.
Stop producing videos first and wondering why they fail. Test the message before you invest in the medium. Static, then video, then scale.
