Case study · May 2026

Testing three video narratives in 14 days.

How CornerCast made apples-to-apples video A/B testing affordable for a 41-year-old American manufacturer — and returned 7.6× on ad spend along the way.

Murphy Bed Company · USA 14-day pilot · Apr–May 2026 YouTube CTV
3
Creatives shipped
tested head-to-head
14 days
Flight time
same audience, same window
0.0×
Return on ad spend
YouTube native attribution
0.0%
Best click-through
Video 2 · the hook winner
02 / THE CLIENT

Patent-protected Murphy beds, built in USA since 1985.

The company has manufactured patent-protected Murphy bed mechanisms in USA since 1985. Their customers — DIY homeowners, cabinet makers, contractors, ADU and tiny-home builders — buy the mechanism for $349 and build a custom Murphy bed for under $1,000.

The alternative is a $5,000–$10,000+ custom commission, or a generic off-the-shelf unit that doesn't match the home.

The product is excellent. The marketing challenge is the one facing thousands of American manufacturers: how do you reach the right buyer when high-quality video has historically consumed a full year of marketing budget per spot?

Founded
1985
Headquarters
USA
Product
Murphy bed mechanism (patent-protected)
Buyer
DIY builders, contractors, ADU / tiny-home
Price point
$349 mechanism
03 / THE CHALLENGE

Video A/B testing has been a luxury — until now.

Marketers have A/B tested email subject lines, landing pages, and display ads for two decades. Video has been the holdout. Producing one high-quality commercial takes weeks across crews, talent, and post — producing three to test against each other has been the exclusive domain of brands with seven-figure media budgets.

The result: most advertisers ship one creative, hope it works, and never see what the alternatives would have done. They pick. They guess. They miss.

Customer didn't want to guess.

04 / THE APPROACH

Three creatives. Two briefs. One campaign.

CornerCast worked from two creative briefs built against distinct mindsets in the Murphy bed buyer journey, then produced three AI commercials to test those narratives head-to-head.
Creative Brief 01

Buy vs. build

Prospects weighing whether to buy an off-the-shelf Murphy bed or build their own.

Creative Brief 02

Comparing mechanisms

Prospects who have already decided to build, and are comparing mechanisms.

Video 01
Narrative pull

Lowest CTR, strongest downstream engagement.

Video 02
Hook winner

Highest click-through rate (3.1%).

Video 03
Steady performer

Middle of the pack on every metric.

All three shipped simultaneously on YouTube against the same target audience over a 14-day window — YouTube's recommended practice for native A/B testing. It gave the customer something they'd never had: three real narratives, evaluated on the same buyers, in the same window.

05 / THE RESULTS

Three creatives, three distinct signals.

The headline finding: the creative with the strongest hook wasn't the one with the strongest commercial result. Exactly the kind of insight no single-creative campaign can produce.
Click-through rate · same audience, 14-day window
Video 11.5%
Video 23.1%Hook winner
Video 32.8%

Hook ≠ conversion

Video 2 won click-through (3.1%). Video 1 — despite the lowest CTR — drove the most overall engagement and the cleanest downstream behaviour (drove more sales).

Retention was a wash

~51% average view duration across all three. Differentiation lived at the top of the funnel — the hook and the audience it attracted — not in whether viewers stuck around.

Genuinely new reach

92–97% new-viewer rate across the board. The campaign expanded reach rather than re-serving an existing audience.

The next move: pair Video 1's narrative pull with Video 2's hook efficiency, then run the hybrid against a fresh cohort.

06 / THE RETURN

Oh, and the campaign returned 7.6×.

The real value of this pilot is the testing infrastructure. The commercial outcome was also strong: the campaign produced 10 confirmed orders — a 7.6× return on ad spend over 14 days.

Return on ad spend is revenue attributed to a campaign divided by the amount spent on it. A 7.6× ROAS means every dollar of ad spend brought back 7.6 times as much in tracked sales.

0.0×
Return on ad spend · 14-day YouTube pilot
Confirmed orders
10
Flight
14 days
Platform
YouTube
07 / WHAT IT UNLOCKS

From "which one?" to "which three?"

For the customer, the pilot delivered three things at once: a working campaign, a measurable return, and three real creative comparisons that will inform every decision in the next cycle.

For the broader market of manufacturers, retailers, and SMBs, the implication is larger. Video A/B testing — table stakes for the rest of digital marketing for twenty years — is finally within reach. The question stops being “which one creative should we make?” and becomes “which three should we test?”

Your turn

Which three creatives would you test?

Tell us about your business and we'll produce your first head-to-head video test — done for you, on premium Connected TV.

Now booking — limited production slots this month.