Mid-six-figure DTC apparel brand with strong organic demand but a paid account that had been stitched together by three different agencies. Creative was stale, the structure was bloated with 40+ ad sets, and the team was making scaling decisions on last-click ROAS in Shopify.

Audit revealed heavy audience overlap, learning phase fatigue across 70% of ad sets, and a creative pipeline producing one new concept per month — not nearly enough to outrun fatigue at the spend level they wanted to hit.

• Consolidated to a 2-campaign prospecting + 1-campaign retargeting structure with broad targeting and CBO.
• Stood up a weekly UGC pipeline with 3 creators, producing 8–12 net-new concepts per month.
• Introduced a contribution-margin scaling rule: any ad set above target mCAC gets +20% budget every 48 hours.
• Replaced last-click reporting with a blended view tied to weekly contribution margin, reviewed every Monday.

Within 90 days, daily spend moved from $1.8K to $6.2K while holding a profitable mCAC. The brand has stayed at that spend level for two consecutive quarters with consistent creative throughput.

Designer furniture studio in Sydney with AOVs between $1.8K and $6K. Buyers typically researched for 3–8 weeks before purchasing, which made campaign signal slow and noisy.

The previous setup optimised for purchase events that arrived too late and in too small a volume to teach the algorithm anything useful. The team was killing campaigns before they had a chance to learn.

• Defined a custom 'qualified session' event (3+ pages, 90+ seconds, key product viewed) and optimised cold campaigns toward it.
• Produced product-education video for each hero SKU — materials, dimensions, designer story.
• Built a 4-stage retargeting ladder mapped to research, comparison, decision and post-visit windows.
• Reviewed performance on a 28-day window with a blended attribution view alongside platform data.

Qualified sessions more than doubled, blended ROAS held at 4.2× across the longer window, and attributed revenue grew 74% year over year without a proportional increase in spend.

Multi-location plumbing company across five cities in the US Sunbelt, heavily dependent on Google Ads for lead flow. Two prior Facebook tests had failed to produce profitable leads.

Earlier tests treated plumbing as a single service with one generic creative. In reality, an emergency burst-pipe customer and a planned water-heater install behave nothing alike — they need different hooks, different offers and different response times.

• Split campaigns into 'emergency' and 'planned service' tracks, each with its own creative and landing experience.
• Built location-specific ad sets per city with locally shot creative and named technicians.
• Added a one-question pre-qualifier on the form (issue type) so the call team could prioritise the queue.
• Set a weekly review with the operations lead to pause underperforming cities and reinvest into top performers.

Cost per lead dropped 41% across the five-city footprint and monthly revenue grew 58% within two quarters. Meta now sits as a stable second channel alongside Google, removing the single-channel risk.