How Global Healing found one of its highest-performing channels with Prescient
Global Healing is a Houston-based health and supplements brand founded in 1998 by Dr. Edward Group. The company sells premium natural supplements, cleanses, and detox protocols through three main storefronts: its own ecommerce site, wholesale, and Amazon, with Amazon picking up a growing share of total revenue. To drive sales D2C, Global Healing runs paid media on nine channels at once: YouTube, Meta, AdWords, Pinterest, Bing, Amazon Ads, Amazon DSP, Impact, and CTV.
The Challenge
Each platform's dashboard told Global Healing a different story about what was working, and none of them were consistent with one another. YouTube was the worst offender because the platform report showed virtually zero revenue attribution. Every dollar spent on YouTube seemed to vanish into the feed, making YouTube appear to be the worst performer among the nine channels in Global Healing's portfolio.
With each passing day, it felt like burning budget for lack of trackable ROAS. For Global Healing, where other tactics like Amazon Ads and Google Search already claimed credit for more than half of attributed sales yet were difficult to scale further, the gap between exposure and final purchase was too painful to ignore. Cutting the YouTube budget based on dashboard data is the decision most growth teams would make. For Global Healing, it would have been a mistake.
The team boiled it down to three questions:
- Which channels are actually driving revenue once cross-channel halo effects are on the table, beyond what each platform claims for itself?
- How confident can we be in moving more budget into a channel that every dashboard shows as underperforming?
- Is it worth continuing to invest in YouTube?
The Solution
The standard approach would have been to cut back on YouTube and redirect the budget to the channels with the highest ROAS. However, the Global Healing team knew that this strategy would be short-sighted, and turned to Prescient's modern Marketing Mix Model for a more holistic view of cross-channel performance.
Prescient analyzed daily historical data across the nine channels, with an Amazon revenue model fit exceeding 90%, and YouTube turned out to be the channel with the highest return in the portfolio once the model accounted for cross-channel halo effects. Notably, YouTube generated revenue for both DTC and Amazon, with nearly two-thirds of it showing up as Amazon purchases through halo effects. No platform dashboard could trace that revenue back to the original YouTube exposure, given the indirect nature of awareness driving search intent elsewhere.
The model also revealed a common gap in Amazon's in-platform reporting. Amazon's reports attributed substantially more revenue to Amazon Ads than the validated model confirmed. In practice, Amazon Ads was taking credit for purchases that YouTube generated. Identifying this pattern shifted the narrative from "which channel should we scale?" to "which channel creates demand and which one capitalizes on it?", turning budget decisions into a more balanced, top-to-bottom funnel flow.
While Global Healing knew YouTube had immense scale potential, the team opted to increase spend gradually, watching for where performance might saturate. The team ran execution through three Prescient products: the MMM attribution platform to monitor model health and performance, the Channel Optimizer to plan budget changes and pinpoint diminishing returns as new data arrived, and the Halo Effects Models to track how value flowed between channels over time.
The Results
View-through channels lead, with YouTube on top
Global Healing not only lifted its ROAS but also gained a comprehensive view of its entire media portfolio looking at Prescient's dashboard. The picture inverted: a channel the platform ranked last across the nine-channel mix turned out to be the one pulling the most weight, once cross-channel halo effects entered the model.
Where YouTube's value actually goes
Prescient helped to see how revenue shifted by destination to get a complete picture. 64% of YouTube's modeled revenue landed as Amazon purchases. 28% landed as direct ecommerce sales. 8% showed up as indirect ecommerce halo. For a brand where Amazon already represents the majority of total business, a channel that reliably drives Amazon purchases is effectively a direct revenue channel. The only reason it looked like brand awareness was that no platform dashboard could connect the exposure to the eventual sale.
Scaled YouTube 8.4x while efficiency improved
With the MMM data in hand, Global Healing scaled YouTube spend in stages, watching the modeled performance offered by Prescient at each step. Testing ran at baseline spend with 49% of YouTube's value arriving on Amazon. The building phase moved to 1.7x baseline with 53% Amazon halo. The scaled phase ran at 8.4x baseline with 67% Amazon halo. Modeled efficiency held at every spend level. At the highest point, return even improved slightly, and the Amazon halo share kept growing with spend. Over the same period, overall weekly revenue grew roughly 21%, and Amazon's share of total business revenue moved from 52% to 62%. YouTube's ranking shifted from #9 of 9 by platform attribution to #1 of 9 by modeled return once spend reached the scaled phase.
Key takeaways: What this means for you
- Platform reporting can be directionally wrong, not just imprecise. YouTube went from last to first in Global Healing's channel ranking when Prescient's MMM measured it. If your business runs across both DTC and a marketplace like Amazon, the channels that look like your worst performers on the dashboard may be driving the majority of your revenue on a different destination. Pulling budget on dashboard signal alone is the fastest way to cut your best channel.
- Amazon halo effects are the largest source of hidden channel value for brands selling on both DTC and Amazon. Nearly two-thirds of YouTube's value appeared as Amazon purchases. A view-through channel like YouTube, Pinterest, or connected TV can show near-zero in-platform ROAS for a simple reason: its purchases land on a storefront the ad platform cannot see. Any portfolio that mixes DTC and marketplace has to treat the halo question as a portfolio-level question.
- Channels that look efficient at low spend may have significant headroom. Global Healing scaled YouTube 8.4x and modeled efficiency improved when they looked at Prescient's panel driven by halo effects. Dashboard logic would have called that reckless. The data behind the MMM said otherwise, and the team trusted the data.
In the past, marketing has largely been plagued by inconsistent and unreliable attribution, leaving teams feeling either inflated or deflated from performance. Today, the expectation for quick, insightful, and (most importantly) accurate data has never been higher. Companies want to understand exactly where their money is working for them, and where it is not.
