The account was stuck at 2.1x ROAS for three months
An ecommerce brand selling a consumer product in the $40 to $80 range. Monthly Meta spend of $12,000. The campaign had been running Advantage+ Shopping with six static image ads, refreshed quarterly. ROAS held steady at 2.1x for three consecutive months. Not catastrophic, but not growing. The operator assumed the product had reached its natural ceiling on Meta.
The audit suggested otherwise. The campaign structure was fine. The targeting was broad (Advantage+ was doing its job). The landing page was decent: mobile-optimized, product-focused, reasonable load time. The bottleneck was the creative.
Six static images, all variations of the same product-on-white-background format. No hooks. No story. No demonstration. No social proof in the ad itself. The ads looked like catalog entries, and they had been running long enough that the audience had seen them multiple times. Frequency was at 4.2. The creative was exhausted.
What changed
We produced eight short-form videos over two weeks. All under 30 seconds. All shot in a native, informal style: product in hand, someone demonstrating it, text overlays with a specific benefit claim, and a clear call to action at the end. Production cost was minimal because the format does not require studio lighting or professional editing. It requires a phone, good light, and a clear script.
The videos were loaded into the existing Advantage+ campaign alongside the static images. No changes to bidding. No changes to the landing page. No changes to budget. The only variable was the creative.
What happened in six weeks
ROAS went from 2.1x to 4.3x. Not gradually. The inflection was visible within the first ten days. The Advantage+ system recognized that the video ads converted at a higher rate, shifted budget toward them, and the blended campaign performance lifted.
The static images did not disappear from the campaign; they continued to serve at a reduced share. But the video ads consumed roughly 70% of impressions within three weeks because the algorithm followed the conversions.
Cost per purchase dropped from $38 to $19. Average order value did not change. The improvement was entirely in conversion rate: users who watched a video ad and clicked through were twice as likely to buy as users who saw a static image and clicked through.
Why video works differently on Meta
The mechanical reason is placement access. Static images serve primarily in Feed. Video serves in Feed, Reels, Stories, and in-stream placements. Reels in particular has expanded significantly as an ad surface, and its inventory is cheaper than Feed because competition among advertisers is still building. An account running only static images is excluded from Reels entirely and competes only in the most expensive surface.
The behavioral reason is attention. A static image competes for a scroll-stopping fraction of a second. A video, when the hook works, holds attention for five, ten, fifteen seconds. That additional time communicates more information: what the product does, how it looks in use, what other people think of it. The user who watches 15 seconds of a product demonstration arrives at the landing page with more purchase intent than the user who glanced at a product photo.
Short-form video can increase purchase likelihood by up to 80% 1. That number aligns with what we saw in this account: the video ads did not just reach more people. They persuaded more of the people they reached.
What did not change
The landing page was the same throughout. The product was the same. The price was the same. The audience was the same (Advantage+ broad targeting). The budget was the same. The bid strategy was the same.
The only thing that changed was what the ad looked like and how long it held attention. That single variable doubled ROAS.
This is the firm's second position expressed in numbers: the leverage now sits in the ad itself. The platforms find the buyers. Your job is to give them something worth watching.
The production model this implies
Ecommerce brands running Meta at $10K or more per month should be producing video creative continuously, not quarterly. The fatigue window on Meta has compressed to 5 to 7 days for high-frequency campaigns 2. That means a winning video format needs two to four variations per week to stay ahead of decay.
The production does not need to be expensive. A phone, a ring light, a 15-second script, and a clear product demonstration. The format that wins on Reels is the format that looks like organic content: informal, direct, human. Overproduced brand films do not outperform a creator holding the product and explaining why they use it.
The investment is in volume and speed, not in production value. The accounts that produce one hero video per quarter are running fatigued creative for 11 out of 13 weeks. The accounts that ship new variants weekly stay in the performance window continuously. Over a quarter, the cumulative difference is the distance between a 2x and a 4x ROAS.