MIRANEST › Winning Ads › 03 · V2004-3
My Pomeranian used to…
A first-person owner story about one black Pom — bald chest to full coat — with Before/After labels near the end.
17 shots, ~2.3s each · female VO, 186 wpm — the fastest of the ten · captions throughout · product (box) first at 0:15 · Before » After labels 0:30–0:35 · “20% OFF · Shop Now” pill at 0:36 · script adapted from “My Chihuahua…” in the N Scripts doc
Where people leave share of impressions still watching · last 30 days
22.2%0:03 · bald chest — “skin got so bad”
7.0%0:10 · “if I could go back in time”
2.9%0:20 · eating the jelly
1.8%0:30 · walking with a friend
1.0%0:39 · “can’t live without them”Where the clicks come from
8 of 14 orders came from Reels (57%) · all ads: 34%
Orders Shopify could trace to a placement. Right-hand % = share for all ads.
338 link clicks · bars = people still watching
Can’t pin it down for this one: 2,696 people are still watching at 0:10 — plenty to produce the 338 clicks at any point in the video.
Shopify sessions carrying this ad’s link. Meta credits 24 orders; Shopify can only trace the ones that came through the link.
Shot by shot frame + voiceover or on-screen text
My Pomeranian used to live with a dry, thinning coat.
His skin got so bad…
…it literally looked like a vet’s office
If I could go back in time…
…just gotten these instead.
We’re on our third box
He’s barely scratching anymore
visibly softer and fuller
These aren’t just masking his symptoms
They work from the inside
I am so grateful for these.
We truly cannot live without them now.
Week by week orders per week, Mon–Sun · green = CPO under the account average
What the numbers say
One dog, start to finish
The same black Pom goes from bald chest to full coat — the Before/After at 0:30 is his own.
The one that didn’t fade
Orders every week since launch. Sep 7–27: 11 orders at ¥12.1k, while the account average went up.
Holds attention
Hold 5.0% vs 4.2% median; hook 22.2% is on the median.
Meta picked this cut
Same audio as V2004-1, but it got ¥393k of spend to -1’s ¥174k — and kept its CPO.
Same concept, other cuts
| Ad | Angle · style | Spend | Orders | CPO | Hook |
|---|---|---|---|---|---|
| V2004-1 | Coat · Review | ¥174k | 12 | ¥14,521 | 26.0% |
| V2004-3 | Coat · Review | ¥393k | 24 | ¥16,381 | 22.2% |
| V2004-5 | Coat · Story | ¥39.1k | 1 | ¥39,135 | 27.3% |
| V2004-8 | Coat · Story | ¥34.2k | 0 | — | 19.1% |
| V2004-2 | Coat · Review | ¥3.7k | 0 | — | — |
| V2004-4 | Coat · Story | ¥2.2k | 0 | — | — |
| V2004-6 | Coat · Story | ¥2k | 0 | — | — |
| V2004-7 | Coat · Solution | ¥0.5k | 1 | ¥498 | — |
Meta data from the HK hub, pulled 2026-09-28 · window Jul 31 – Sep 28 · medians = the 26 ads that spent ¥50k+; video medians = the 16 videos with 5k+ impressions · hook and hold are Meta’s 30-day video fields · voiceover transcribed from the file and checked against the N Scripts doc · original on Drive · full notes: Projects - Omega/Winning Ads Breakdown/V2004-3/V2004-3-breakdown.md