YMP
Ecommerce · Food · Meta Ads Case Study

545% Growth In 5 Months

Uncle Butch Fudge had product-market fit, a loyal repeat base, and a Meta account stuck running the same product-led creative to every buyer regardless of why they were actually purchasing. Over a five-month window we restructured the account around multiple buyer angles, engineered the offer for both gifting and self-purchase, and built a UGC engine that fuelled creative across all of them. Revenue grew 6.5x at a 2.97x ROAS.

business.facebook.com · Ads Manager · Confidential
545% Growth In 5 Months, account results
6.5xRevenue Growth
$630kRevenue (5 Months)
5moPeriod
2.97xAccount ROAS
TL;DR
  • Stopped speaking to "fudge buyers" as one audience. Ran multiple parallel angles, gifting, self-treat / comfort food, family-sharing, subscription/repeat, and occasion-specific, each with its own creative DNA and audience signal.
  • Engineered the offer to serve both buyer modes: bundle SKUs and multi-recipient checkout for gifters, single-tin and subscription anchoring for self-purchasers.
  • Treated Nov–Mar as one continuous campaign window, not three holiday launches. No cold-starts between Thanksgiving, Christmas, and Valentine's, so learnings and audience signal compounded.
  • Built a UGC engine producing multiple content types (reaction, taste-test, "treat night," family moments) so every angle had fresh creative without falling back on studio product shots.
The Challenge

One Creative Voice For Five Different Buyers.

The brand made genuinely loved fudge, but the account was treating every prospect as the same person. The audience wasn't one buyer, it was several. Someone buying as a gift for their mum needed a completely different message from someone buying a tin for themselves on a Sunday night, and both were different again from the subscriber, the family-sharing buyer, and the "I need something for the office before Friday" gifter. None of those differences were showing up in the ads.

Product-Led Creative For An Angle-Driven Category

Every active concept led with the product itself: glossy studio shots, ingredient callouts, slow-pan packaging hero shots. Creative was on-brand and beautifully shot, and completely unable to differentiate between the five different reasons someone might buy. Without angle differentiation, the algorithm was averaging across a wide intent spread and the creative was speaking to none of them well.

AOV Stuck, For Both Buyer Modes

The catalogue and checkout flow had been built around one assumption: a single buyer purchasing one tin for one address. That left money on the table on both sides. Gifters with multiple recipients couldn't easily send to multiple addresses in one order. Self-purchasers had no subscription anchor, every reorder required a full repeat checkout. Average order was effectively capped by the structure of the offer, not the willingness of the buyer.

Three Holiday Launches Instead Of One Continuous Window

Thanksgiving / Black Friday, Christmas, and Valentine's were being run as three separate launches, spun up two weeks before, ramped, then turned off. Every gap was a cold-start. Every launch had to re-learn audiences the account had already paid to teach the algorithm about six weeks earlier. The single most purchase-dense window of the year was being run as three disconnected sprints.

Studio Photography In A Category That Sells On Real Moments

Whether the angle was gifting, self-treat, or family sharing, the thing that actually makes someone stop scrolling on a food ad is people, eating it, reacting to it, sharing it. The account was leaning entirely on studio product photography. None of the human, in-context, real-life content that food brands win on was in the rotation.

The Approach

Build For Multiple Buyer Mindsets In Parallel, Not One Hero Angle.

The shift wasn't picking a single message. It was running multiple angles intentionally, in parallel, each with its own creative, its own audience read, and its own contribution to the window. Once the account was set up to serve all of them, the algorithm had something to lean into for whichever buyer mindset was showing up in any given auction.

01

Multi-Angle Creative Architecture

Replaced the single product-led creative track with five running in parallel:

  • Gifting, recipient-led framing ("the one gift my brother actually said thank you for"), multi-recipient hero shots, gift-message close-ups
  • Self-treat / comfort food, Sunday-night, "you earned this," solo-indulgence framing
  • Family / sharing, movie night, "the kids fight over this one," kitchen-table moments
  • Subscription / repeat, "the tin that arrives before you run out," habit-based framing for the regular buyer
  • Occasion-specific, keyed to Thanksgiving, Christmas, Valentine's, with creative briefs refreshed 3–4 weeks ahead of each moment

Each angle ran with its own hook variations and its own audience signal feeding back into the algorithm. The wins across angles were genuinely different concepts, gifting winners didn't carry self-treat performance and vice versa, which is why running them in parallel was the unlock.

Multi-Angle Creative Strategy
02

Dual-Mode Offer Engineering

Built the offer to serve both buyer modes:

  • For gifters: bundle SKUs (3-tin, 6-tin, sampler), "send to multiple addresses" flow, gift-message presets, gifting-led PDP variations
  • For self-purchasers: subscription anchoring on the PDP, single-tin pricing held accessible, "Subscribe & Save" framing with skip-anytime flexibility
  • Cross-mode: post-purchase upsells tuned to which mode the customer bought in, gifters got "add to the gift list" suggestions, self-purchasers got subscription conversion offers

AOV climbed materially across the window because the offer architecture finally matched the actual mix of buyers.

Dual-Mode Offer Engineering
03

Continuous Window Campaign Architecture

Killed the three-launches-a-year approach. Built one continuous campaign structure running from late October through Valentine's, with creative swaps timed to each moment but the same campaigns, ad sets, and audience signal carrying through. No cold-starts. No learning resets. By February the algorithm had four months of compounded purchase data feeding into Valentine's, instead of treating it as a fresh launch.

Continuous Window Architecture
04

UGC Engine Feeding All Five Angles

Seeded the existing customer base with simple, angle-specific briefs:

  • Gifters: film the recipient opening the box
  • Self-treat buyers: film your "treat moment," tea, blanket, first bite
  • Family buyers: film the family fighting over the last piece
  • Subscribers: film your tin arriving and being added to the stash

Built a steady weekly intake of raw, vertical, phone-shot content covering every angle. Polished brand assets were used in Feed; UGC carried Reels, Stories, and short-form. The variety of UGC types meant no single angle ran out of fresh creative across a five-month window, which is usually what kills creator-light brands in long-running campaigns.

Angle-Specific UGC Engine
Limited Spots. 3 Onboarding Slots Left This Month

Ready to stop guessing?

Apply for a free Paid Ads Audit + Custom Scaling Plan Powered By Our Predictable Scaling Quadrant™ Method. We'll pinpoint exactly where your budget is leaking and show you a clear path to profitable scale.

Full audit of your current acquisition setup
Custom growth roadmap for your offer
Honest breakdown of what will & won't work

Joined by 1,500+ businesses