Clip vs still, slot by slot

Every media slot on the 6-reasons listicle, what's in it now, and what else we hold. Clips autoplay muted and loop, same as they would on the page. Sources: Michael's 2026-08-11 facility b-roll (13 silent, caption-free cuts), the original demo footage, and 103 stills pulled from the Motion ad account.

Clip wins Still is right Needs a decision first

Hero

Top of page, beside the headline Still is right

The hero has to be readable in under a second and it is carrying a labelled comparison. A clip would make the reader wait for the payoff that the annotated still delivers instantly. Keep the still. The hot-water clip belongs further down, where the reader has already been told what to look for.

Annotated hot-water comparisonIn use

Annotated two-glass comparison

NO CREATINE / CREATINE labels burned in. The whole argument in one frame.

1:1 nativeBurned-in labels
Clip option

Full hot-water demo

The original ad footage, start to finish. Has spoken-caption subtitles burned in.

7.4 MBSubtitlesVertical
Two empty glasses set up for the testStill option

Test setup, both bags

Before shot. Reads as "here's the experiment" rather than "here's the result".

Competitor pack in frame

1. Most creatine gummies barely have creatine in them

Reason 1 Clip wins

This is the one slot where motion does real work. The claim is that one glass changes and the other doesn't, and a still can only assert that, a clip proves it. The reader watches sediment fall out of one glass while the other stays clear. Cost is the subtitles burned into the source footage and the file weight.

Two glasses, one with sedimentIn use

Two glasses, captions cleaned

Result frame with the ad subtitle removed. Static.

No burned textCrops to 1:1
Recommended

Hot-water test, full demo

The sediment actually settling. Needs a trim to the dissolve beat and a caption decision.

7.4 MB rawSubtitlesReal proof
Bowl held up showing sedimentClip option

Bowl from underneath (WebP)

Already converted, 628 KB. Different demo: a bowl held up to the lens.

628 KBSubtitle
Competitor glass, clear waterStill option

Competitor glass, nothing settled

The negative half on its own. Pairs with the Crave glass as a 2-up.

Competitor pack in frame

2. The reason is heat

Reason 2 Still is right

There is no footage of creatine degrading, and there never will be — it is a chemical fact, not an event you can film. The built graphic is the right answer here and it is the only slot on the page carrying a number the reader has to hold on to. Any clip we drop in would be decoration standing next to the argument rather than making it.

Crave▼ Most gummies▼
150°F

Built temperature scale

CSS, no asset. Sharp at any size, changes in seconds.

0 KBEditable
Clip option

Amber base in the kettle

Could sit here as "this is the cook that kills it", but it repeats reason 3's visual.

LoopsNo peopleDuplicates R3

3. So Crave cooks it backwards

Reason 3 Clip wins

Already a clip, and it should stay one — this is a process claim and process is the thing video is for. The current cut is a two-beat edit of clips 04 and 07 (base cooking, then powder going in), which matches the two-beat copy. Worth considering: the copy is now four numbered steps, so a longer cut could carry all four. The depositing clips below are cut and ready and currently have no slot anywhere.

Precook processIn use

Two-beat precook cut (WebP)

Base cooked, then creatine folded in. Converted from GIF: 5.6 MB down to 972 KB.

972 KBNo captions
Beat 1

04 · Amber base mixing

5.8s. The pot before creatine goes anywhere near it.

LoopsNo people
Beat 2

07 · Creatine into the mixer

3.2s. Literally the moment the copy describes.

No facesGloved hands
Unused

11 · Depositor filling molds

7.9s. Reads as a real plant. No slot on any page yet.

LoopsNo people
Unused

08 · Pink mixture mixing

6.7s. The batch after creatine is in. Could close a four-step cut.

LoopsNo people
Unused

12 · Racks of finished gummies

4.9s. Pans out to scale. Good final beat.

No people

4. Every batch gets sent to an independent lab

Reason 4 Still is right

A certificate is a document and a reader wants to stop and look at it. Motion actively hurts here: a moving document can't be read, and every figure that passes through frame becomes a claim we have to source. The one lab clip we hold is flagged for exactly that reason.

Full Eurofins certificateIn use

Complete Eurofins certificate

One whole page, signed, Page 1 of 1. Result reads 264,000 µg/g.

One figure onlyAd caption burned in
Needs a call

06 · Reports over the facility

2.1s. Multiple ANALYTICAL REPORT pages sweep through frame.

Unsourced figures in frameUnreadable in motion
Fanned report pagesStill option

Four reports fanned

Sells "every batch" better than one page. But shows 246,000 µg/g, which clashes with the 264,000 above.

Conflicting figure

5. You can test it yourself in 90 seconds

Reason 5 Clip wins

The copy gives an instruction — drop three in, stir, wait. An instruction is a sequence, and a still can only show its last frame. This is the second-strongest case for motion on the page after reason 1, and the same source footage covers both, so one trim decision serves both slots.

Glass bottom with sedimentIn use

Glass base, sediment settled

Answers the "can't see the bottom" note. Shows the result, not the method.

SubtitleBase fills frame
Recommended

Hot-water test, full demo

Trim to the stir-and-settle beat. Same source as reason 1, different section of it.

7.4 MB rawSubtitles
Gummies poured into a palmStill option

Three gummies into a palm

Step one of the instruction rather than the result. Reads as "start here".

No competitor pack
Cups with gummies soakingStill option

Cups mid-soak

The middle of the test. Less finished-looking, more believable as a home demo.

Check framing

6. And they actually taste like candy

Reason 6 Worth testing both

You called the current photo great and I'd keep it as the default. But taste is an appetite argument, and appetite responds to motion — sugar catching light, gummies tumbling. If any slot other than the proof sections deserves a clip, it's this one. The catch: the only gummy-handling clips have a face in frame (02) or are finished-product b-roll rather than food styling.

Stacked gummies with fruitIn use

Stacked gummies, hand held

Your pick. Sugar coating and colour both read instantly.

1:1 nativeNo text
Clip option

Gummy pour

2.8 MB. Gummies tumbling out of the bag. Closest thing we have to an appetite shot.

2.8 MBCheck for captions
Clip option

12 · Racks of green gummies

4.9s. Macro on sugar coating, then pulls out. Clean, no people.

No peopleReads factory, not kitchen
Macro of a sugar-coated gummyStill option

Sugar-coating macro

Tightest texture shot in the set. Strong appetite still.

No text

Offer box

Bottom of page, beside the CTA Still is right

The job of this image is to show what arrives. A reader at the buy button is deciding, not being entertained, and a looping clip beside a CTA pulls the eye off the button. Keep it static. If anything, swap for a pack shot that shows all three flavours so the flavour bullet has something to point at.

Handful of gummies from the bagIn use

Handful poured from the bag

Moody, dark. Product-in-use rather than product-as-sold.

1:1 native
Three bags and three piles of gummiesStill option

Three bags, three piles

Shows the full range. Matches the flavour bullet directly.

Clean studio
Open pouch of green gummies, overheadStill option

Open pouch, overhead

What it looks like when you open it. Good "what arrives" frame.

No text

The short version. Three slots earn a clip: reason 1 and reason 5, because both describe something changing over time, and reason 3, which already has one. Everything else is better static — the hero has to read instantly, reason 2 is a fact rather than an event, reason 4 is a document you want to stop and read, and the offer image sits next to the button and shouldn't compete with it. Reason 6 is the genuine coin-flip.

What's blocking the two hot-water slots. Both want the same source file and it has spoken-word subtitles burned into it. Either we re-cut around the subtitled stretches, re-burn captions we actually want, or accept the ad's own captions on a landing page. That's a call, not a task. The file is also 7.4 MB raw and would need trimming and converting before it goes anywhere near a live page — for reference, the process clip came down from 5.6 MB to 972 KB.

Worth knowing. The facility b-roll has six cuts with no slot on any current page, including two depositor shots that are the strongest "this is a real plant" footage we hold. Also flagged in the original placement notes: clips 02 and 03 have identifiable faces, so confirm Crave has releases before either goes on a public page.