Skip to content
Campaign Structure

Full-Funnel Paid Strategy

Coordinating paid campaigns across prospecting, retargeting, and retention at once, instead of optimizing only for bottom-funnel conversions.

Your retargeting campaigns show the best ROAS in the account, so the budget drifts toward them. Warm audiences convert, the number looks great, and the top-of-funnel line item keeps getting trimmed. A few months later the retargeting pool is shrinking, frequency is climbing, and blended performance is sliding even though every individual campaign still reports a healthy return.

A full-funnel paid strategy is the attempt to prevent that. It runs prospecting (finding new buyers), retargeting (converting people who already engaged), and retention (winning back existing customers) at the same time, so the algorithm has upper-funnel signal to work with and the warm pool keeps getting refilled.

That is the framework. What the teardowns show is that D2C brands rarely run it the way the framework describes.

Brands split the funnel across platforms, not inside one

The published playbooks assume one platform carries all three stages. In the d2cplay corpus, the more common pattern is that each platform is assigned a single job, and the funnel is stitched together across two of them.

Ridge Wallet points every one of its 273 Meta ads at a purchase, with a “Shop Now” CTA on all of them and no top-of-funnel “get to know us” campaign anywhere in the account, in that teardown. Awareness is handled off Meta, by organic and influencer channels. Meta is the conversion machine, not the funnel.

Fresh Clean Threads runs the same shape. All 150 of its Meta ads carried a “Shop Now” CTA as of March 15, 2026, only 6 were retargeting units, and the teardown’s read is blunt: FCT does no top-of-funnel education on Meta. Its Google account, meanwhile, carried zero customer-list targeting — no remarketing there at all. One platform sells, the other prospects.

When Google is the only funnel there is

The cross-platform split assumes both platforms are running. In markets where only one is, the “funnel” collapses to a single stage.

Our Place runs Google Search in 42 countries but concentrates its Meta and Instagram spend on a handful. A shopper in Poland, Greece, or the Czech Republic never gets a Facebook retargeting ad — their only exposure to the brand is the Google search result when they type “non-toxic cookware.” There is no awareness stage and no retargeting stage for that buyer. The search ad is the entire funnel.

MUD\WTR shows the intended cross-platform version working the other way. The teardown documents Google display ads driving cheap traffic to educational content, with the reader then retargeted on Meta — Google as the top of the funnel, Meta as the bottom. The two platforms only add up to a full funnel when both are deliberately assigned a stage.

The budget-split rules disagree, and the accounts follow none of them

The reason a fixed split is hard to apply is that the published splits do not agree with each other. Keeping prospecting and retargeting in structurally separate ad sets delivers 15–25% lower CPA, per Meta Marketing Agency’s DTC playbook, and brands combining Search, Performance Max, and Meta into one system beat single-channel setups by 31% on blended ROAS, per Adscrey’s 2026 performance-marketing report. Neither tells you the ratio.

Where a ratio does get named, it moves. AdYogi puts 30% of budget into top-of-funnel at $50K/month spend and 20% at $150K/month, on the logic that the split is a stage-of-growth question, not a setting. A brand large enough already has a retargeting pool big enough to sustain conversion volume without as much cold prospecting.

None of the corpus accounts read as if they started from a percentage. Ridge and FCT put roughly nothing into prospecting on Meta; Our Place puts everything into a single search stage across most of its markets.

The retargeting trap the platform docs skip

The failure mode that pulls budget toward the bottom of the funnel is a measurement artifact. Retargeting reports a high ROAS because it is credited for converting people who were already going to convert, which makes the warm audience look like the efficient place to spend.

Overweighting it decays the thing it depends on. Custom audiences shrink without fresh prospecting, frequency against the remaining pool rises, and marketing efficiency ratio degrades even while the retargeting campaign’s own ROAS stays high, per AdYogi’s account of the pattern. The individual campaign number and the account-level number point in opposite directions.

This is why blended ROAS and MER exist as separate metrics from platform-reported return. A full-funnel strategy is, in practice, a bet that the blended number matters more than the per-campaign one.

Where the funnel actually gets separated in these accounts

When brands do build explicit stage separation, it tends to be small and surgical rather than a three-way budget split.

RYZE’s Google account contains exactly one ad that excludes existing customers from prospecting and one that remarkets to them — a deliberate prospecting-versus-retention boundary drawn in two ads, not across the account. Little Sleepies keeps its retargeting layer to 22 dynamic product ads, 86% of them launched in a single wave on March 3, 2026, sitting on top of a much larger prospecting account. Magic Spoon runs the clearest top-of-funnel layer in the set: 45 of its 66 Google video ads are brand-level awareness with no product in the headline, lines like “Too old for cereal? We think not,” rather than conversion creative.

The stages are real. They are just rarely balanced to a formula.

Seen in these teardowns

Where we've analyzed Full-Funnel Paid Strategy

Fresh Clean Threads Runs 150 Meta Ads. Only 17 Pieces of Copy Power All of Them.

Meta AdsFull TeardownAds StrategyDTCMar 16 · 14 min read

150 active ads, 87% algorithm-driven, 17 unique copy blocks, and three simultaneous promo codes. Inside a $60M basics brand that trusts Meta's machine more than its own creative team.

Little Sleepies Runs 350 Meta Ads With One Template. One.

Meta AdsFull TeardownAds StrategyDTCMar 23 · 16 min read

350 active Meta ads. 288 of them - 82.3% - are DCO. And every single one uses the exact same template. Inside how a $200M children's apparel brand runs its entire paid operation through one dynamic template and 927 cards.

Magic Spoon Has 109 Google Ads Spanning a Full Year. Protein Pastries Don't Exist on This Platform Yet.

Google AdsFull TeardownAds StrategyApr 8 · 14 min read

A deep dive into Magic Spoon's Google Ads account, exploring their legacy product rotation, Amazon strategy, and unique health priority.

MUD\WTR Has 48 Meta Ads. It Has 480 on Google. And 19 of Them Are About ASMR.

Google AdsFull TeardownAds StrategyApr 16 · 12 min read

A deep dive into MUD\WTR's Google Ads account, exploring their strategy of promoting blog posts as display ads.

Our Place Runs 5 Countries on Meta. 42 on Google. One Ad Has Been Live for 3 Years.

Google AdsFull TeardownAds StrategyDTCApr 26 · 15 min read

703 Google ads across 42 countries. One ad running 1,103 days. Meta median age: 23 days. Google median in Europe: 650-750 days. Same brand, same products, two completely inverted playbooks.

Ridge Wallet Marketing Strategy: 273 Meta Ads, 50 Instagram Posts and Website Analyzed, Full Funnel Breakdown

Meta AdsInstagramFull TeardownFeb 20 · 18 min read

I went through Ridge Wallet's entire Meta Ad Library - all 273 active creatives - and analyzed their Instagram, tech stack, and email flows. 88% of their ads lead with value, not discounts.

Same Brand, Opposite Strategy. I Pulled 400 More RYZE Ads on Google.

Google AdsFull TeardownAds StrategyDTCMar 10 · 14 min read

RYZE runs a completely different playbook on Google - 5 products instead of 1, 40% off instead of 25%, 86% video, and Halloween sale ads still showing in March. 400 ads pulled from Google's Transparency Center.

See also