How Dr. Squatch Used Incrementality to Cut Wasted Ad Spend

A $100M+ DTC brand discovered their ads were taking credit for organic sales

Dr. Squatch grew from a garage operation to over $100M in annual revenue primarily through direct-response advertising on Meta and YouTube. Their viral "cold open" ad format became a DTC marketing case study. By 2023, they were spending millions per month on paid media with platform-reported ROAS numbers that looked strong.

Then they started asking a harder question: how much of that revenue would happen without the ads?

The answer changed their entire media strategy.

The problem: platform ROAS looked great but margins were shrinking

Like many scaled DTC brands, Dr. Squatch hit a point where increasing ad spend didn't produce proportional revenue growth. Their Meta-reported ROAS was 4-5x. Google showed 6-8x on branded search. Retargeting showed 10-12x. By platform math, every channel was printing money.

But actual profit margins were declining. Customer acquisition costs were rising. And when they compared total attributed revenue across all platforms to actual revenue in Shopify, the overlap was significant -- platforms were collectively claiming 1.7x more conversions than actually occurred.

This is the standard DTC scaling trap. Every platform takes credit for every conversion it touches, and at scale, nearly every customer touches multiple platforms. The sum of platform-reported conversions always exceeds actual conversions, and the gap widens as you spend more.

The approach: systematic incrementality testing by channel

Dr. Squatch worked with their measurement partners to run a series of controlled experiments over six months. Each test followed the same methodology: randomized holdout groups suppressed from seeing ads for a specific channel, with conversion rates compared to the treatment group.

Test 1: Retargeting

Platform-reported ROAS: 11.3x

The retargeting test produced the most striking result. With a 15% holdout over 4 weeks, the holdout group converted at nearly the same rate as the treatment group. The incremental lift was just 8%.

Incremental ROAS: 0.9x

This meant their retargeting campaigns were spending millions to reach customers who were already going to buy. The ads weren't creating demand -- they were capturing demand that existed independently and taking credit for it.

For every $1 of retargeting spend, they were getting back $0.90 in truly incremental revenue. The platform said $11.30.

Test 2: Branded search

Platform-reported ROAS: 7.8x

Branded search showed similar results. Users searching "Dr. Squatch" were already aware of and interested in the brand. The test revealed that 82% of branded search conversions would have happened through organic search results.

Incremental ROAS: 1.4x

Not completely wasteful -- there was some incremental value from competitive defense and guiding users to specific landing pages -- but nowhere near the 7.8x the platform reported.

Test 3: Meta prospecting

Platform-reported ROAS: 3.8x

This was the channel leadership was most nervous about testing. Meta prospecting was their primary growth engine, and the fear was that an incrementality test would undermine confidence in their largest spend category.

The results were actually encouraging.

Incremental ROAS: 2.1x

Meta prospecting was genuinely incremental -- the holdout group converted at a significantly lower rate than the treatment group. The 2.1x incremental ROAS was lower than the 3.8x platform claim, but it was well above their profitability threshold of 1.5x.

Test 4: YouTube

Platform-reported ROAS: 1.4x

YouTube had the weakest platform-reported numbers, which made it a frequent target for budget cuts in quarterly planning. Leadership had nearly paused YouTube twice in the previous year.

Incremental ROAS: 1.6x

YouTube was actually more incremental than its platform reporting suggested. The channel drove genuine awareness that led to conversions through other channels -- conversions that were attributed to Meta and Google Search in platform reports but were actually initiated by YouTube exposure.

The reallocation: following the incremental data

Armed with channel-level incrementality data, Dr. Squatch made significant budget changes:

Retargeting: Cut by 60%. They maintained a minimal retargeting presence (abandoned cart sequences and 3-day site visitor campaigns) but eliminated the broad retargeting campaigns that were reaching already-decided buyers. Savings: estimated $1.2M annually.

Branded search: Cut by 45%. They reduced branded search bids and budgets, maintaining presence only on high-competition terms where competitors were actively bidding. They also found they could lower bids significantly without losing click share, since their organic listing appeared directly below the paid result.

Meta prospecting: Increased by 25%. With proven incrementality and headroom below the diminishing returns threshold, they scaled prospecting with confidence. They also shifted some budget from broad prospecting to lookalike audiences based on high-LTV customers.

YouTube: Increased by 35%. The incrementality data justified budget that platform ROAS couldn't. YouTube became a strategic upper-funnel investment rather than a budget-cutting target.

The results: profitable growth instead of wasteful scaling

Six months after the reallocation:

• Total ad spend decreased by 12% (net of increases and decreases)
• Revenue increased by 28% year-over-year
• Blended incremental ROAS improved from 1.4x to 2.3x
• Customer acquisition costs dropped 18% because money moved from capturing existing demand to creating new demand

The counterintuitive lesson: spending less on retargeting and branded search didn't reduce revenue. Those customers converted anyway. The freed-up budget went to Meta prospecting and YouTube, which reached new customers who wouldn't have purchased otherwise.

What this means for other DTC brands

Dr. Squatch's experience follows a pattern we see repeatedly across DTC brands at scale:

Retargeting incrementality is almost always lower than reported. Across 30+ DTC brands we've analyzed, retargeting's incremental ROAS averages 25-35% of platform-reported ROAS. The larger and more well-known the brand, the lower the incrementality.

Prospecting channels are usually under-valued. Attribution models favor last-touch channels. The channels that create initial awareness -- prospecting, video, influencer -- get under-credited relative to their causal impact.

The "always on" channels are the ones to test first. If you've been running a campaign for 12+ months without ever questioning it, that's where incrementality testing pays off most. The campaigns nobody questions are the ones most likely hiding waste.

Brand growth reduces channel incrementality over time. A channel that was highly incremental when you were unknown may have declining incrementality now that you're established. As organic demand grows, a larger share of ad-attributed conversions are organic conversions wearing an attribution label.

Frequently Asked Questions

Did Dr. Squatch's retargeting revenue actually drop when they cut spend?

Total retargeting-attributed revenue dropped by roughly 55%. But total company revenue -- measured in Shopify, not ad platforms -- actually increased because the customers who would have converted organically still converted (through direct visits, organic search, or email). Meanwhile, the reallocated budget to prospecting and YouTube brought in genuinely new customers. The net effect was positive revenue growth with lower total spend. This pattern is consistent across DTC brands that cut retargeting: the "lost" retargeting conversions reappear as organic conversions.

How long did it take to run all four incrementality tests?

The full testing program took six months. Each individual test ran for 3-4 weeks, with 2-week analysis periods between tests. They tested one channel at a time to avoid cross-test contamination. Some brands try to accelerate by testing multiple channels simultaneously using geographic holdouts, which can work if the markets are properly segmented. But for a first round of testing, sequential single-channel tests produce cleaner, more actionable results.

Can brands smaller than Dr. Squatch benefit from incrementality testing?

Yes, but the methodology needs to be adapted. Dr. Squatch's high conversion volume allowed them to run user-level holdout tests with statistical confidence in 3-4 weeks. Brands with lower conversion volumes (under 500/month) may need geographic holdouts, longer test durations (6-8 weeks), or larger holdout percentages (15-20%). The key threshold is having enough conversions in the holdout group -- at least 100 -- for the results to be meaningful. Brands spending $30K+/month on a single channel typically have enough volume for a viable test.

---

Go Funnel uses server-side tracking and multi-touch attribution to show you which ads actually drive revenue. Book a call to see your real numbers.