Magic vs Machine Case Study:
Scaled Huha from $1M to $30M Bootstrapped
With click-based attribution from Wicked Reports, Magic vs Machine agency unlocked Pinterest as a top of funnel winner and scaled Meta spend 3x year-over-year, profitably.
The Brand & The Agency
Magic vs Machine, led by founder Angelo Dodaro, is a DTC growth partner for some of the fastest-growing brands in Canada and the US. They specialize in deep, performance-driven Meta advertising and bootstrapped growth.
Huha is a women’s underwear brand built around a proprietary lining that supports women’s health—combining comfort, design, and function. When Magic vs Machine began working with Huha, the brand had done roughly $1M USD in annual sales after its first year.
Within three years together, Huha has grown to roughly $23M YTD and is on track to reach ~$30M in annual revenue, with 48% repeat purchase rate that holds at scale.
And Wicked Reports is the attribution engine they trust under the hood.
The Challenge:
Scale New Customers Without Losing the Plot
Huha is a paid media–driven brand. Growth is directly tied to how efficiently Magic vs Machine can acquire new customers.
AS SPEND RAMPED:
Platform bias & window games
-
Meta, TikTok, Pinterest each show their own optimistic version of reality.
-
View-through-heavy reporting made channels look better than they truly were.
No clear view of new vs repeat
-
Platforms counted lots of sales… but Huha’s growth depended on first-time buyers, not just happy repeat customers.
-
Advantage+ Shopping looked great in-platform, but was quietly pulling spend into retargeting instead of top-of-funnel.
Black box performance at the top of the funnel
-
Pinterest looked “okay” in-platform.
-
Most of the reported conversions were view-through, with limited click clarity.
-
Angelo suspected value, but couldn’t justify serious budget without hard, click-based proof.
iOS 14.5 and the post-privacy world
-
Pixel-based tracking alone was no longer reliable.
-
Google Analytics wasn’t giving a usable cross-channel answer.
-
Other attribution tools felt oversimplified or even misleading, especially for an omnichannel, fast-scaling brand.
Angelo needed scientific, click-based, omnichannel attribution that aligned with how serious performance marketers actually think.
Why Magic vs Machine Chose Wicked Reports
When iOS 14.5 hit, Angelo was already familiar with Wicked Reports’ reputation and Scott’s thinking on attribution.
WHAT STOOD OUT:
Click-based, first-party attribution
Wicked uses real order IDs, CRM IDs, and time-stamped click history, instead of blending in view-through guesses.
New vs repeat baked into the model
Huha could finally see new customers vs repeat customers by channel, campaign, and creative—not just aggregate revenue.
Deep methodology instead of “pretty dashboards”
Other tools tried to oversimplify attribution into a single number. Wicked embraced the reality:
-
Different windows
-
Different models (first click, last click, linear, LTV views)
-
Different questions for different decisions
Future revenue & LTV views
The team could see revenue that came in weeks or months after the initial click, broken down by cohorts (30/60/90 days and beyond).
New Visitor % & Top-of-Funnel clarity
Wicked surfaced new visitor % by campaign, making it obvious when a strategy was really driving new eyeballs… or just circling the same audience.
For Angelo, Wicked became the “duck detector”: when something in-platform “quacked like a duck” but didn’t quite add up, Wicked could prove whether it was real or just attribution noise.
Implementation:
Building a New-Customer Growth Engine
Magic vs Machine rolled out Wicked Reports across Huha’s stack:
Key Insights Revealed
With Wicked's data foundations in place, Magic vs Machine uncovered two critical insights that changed Huha's growth trajectory.
Pinterest Was a Silent Top-of-Funnel Winner
In-platform, Pinterest looked "fine"—decent ROAS but heavy view-through. Wicked's click-only, delayed revenue lens told a different story.
Without Wicked, Pinterest would've been a "modest side channel." With Wicked, it became a validated discovery engine.
New Visitor % & Advantage+ Shopping
Magic vs Machine suspected Advantage+ Shopping Campaigns (ASC) were over-credited and retargeting-heavy. Wicked made that visible.
Whenever ASC was pushed hard, new visitor % would drop, even when in-platform ROAS looked strong.
This confirmed a dangerous pattern: great-looking ROAS but poor new-customer acquisition.
Using Wicked, Angelo could push back on platform reps and align strategy with long-term growth, not recycled buyers.
Lifetime Value:
How Huha Scales Safely
Huha’s 48% repeat purchase rate is a huge growth lever—but only if you can see it clearly and act on it. With Wicked, Magic vs Machine:
-
Tracks LTV in 30/60/90-day slices instead of waiting for a full “lifetime” to play out.
-
Sees how new cohorts acquired before events like Black Friday behave over time.
-
Uses that insight to decide when to:
- Push harder on acquisition (even if short-term ROAS looks average).
- Lean into launches, new colors, and styles that attract more new customers, not just repeat-only hype.
Results
Revenue growth:
From ~$1M USD to $23M+ YTD, on track for ~$30M.
Repeat purchase rate:
48% repeat rate, holding even as the brand scales aggressively.
Paid media efficiency at
scale:
2–3x year-over-year increases in Meta ad spend with ROAS improving, not declining.
Clear, defensible view of performance across Meta, Google, TikTok, and Pinterest.
Pinterest unlocked as a
viable TOF channel:
3x ROAS over a 120-day linear model.
10x+ ROAS in full LTV impact views for some cohorts.
Validated as a profitable, scalable discovery channel — not just a vanity add-on.
Decision clarity for the
agency & client:
No more gut-based debates with clients or platform reps.
Wicked’s data “puts arguments to bed” and allows both sides to make bold, confident decisions.
How Agencies Can Use Wicked the Way Magic vs Machine Does
