A collection of DTC brands I've helped grow through Meta advertising — with data, creative strategy, and relentless optimization.
When we started working with an online furniture brand in Poland, the team was already running Meta campaigns actively. Traffic wasn't the issue — but that traffic wasn't converting into sustainable, profitable growth. The brand had been running ads for some time, had an established product catalog, and was generating a steady stream of website visitors. On the surface, the account looked healthy.
But when we looked deeper, the real problem emerged: the brand had hit a growth ceiling it couldn't break through on its own. The biggest challenge was scalability. Every time the ad budget was increased, performance dropped rapidly — CPA spiked and ROAS became unstable. The brand was essentially trapped: it could maintain a certain level of performance, but it couldn't grow beyond it without destroying efficiency.
Our initial audit identified three core problem areas that were working against each other: a fragmented campaign structure that starved the algorithm of data, a creative strategy built around static product shots that didn't match how furniture buyers actually make decisions, and a funnel that had no clear top-of-funnel demand generation layer.
| Metric | Before (Baseline) | Issue |
|---|---|---|
| ROAS | 1.6× | Below target, unstable |
| CPA | High, scaling further increased it | Not scalable |
| Creative mix | ~90% static catalog | Low engagement |
| Campaign structure | Fragmented, many small ad sets | Algorithm can't learn |
The first step was restructuring the account. The existing setup had too many small, fragmented ad sets — each collecting insufficient data to let Meta's algorithm learn properly. We consolidated these into fewer, broader campaigns with higher data density. This single structural change alone improved signal quality significantly and allowed the algorithm to start optimizing toward outcomes rather than just delivery.
The second — and most critical — step was rebuilding the creative strategy from the ground up. Furniture is not an impulse-buy category. Users spend days, sometimes weeks, researching before committing. They don't just want to see a product — they want to see it living in a space, solving a problem, fitting a lifestyle. The existing creative mix (roughly 90% static catalog images) was answering the wrong question. Instead of 'what does the product look like?', the creative needed to answer 'how will this change my home?' Based on this insight, we rebuilt the entire creative approach:
On the funnel side, we introduced a proper three-tier architecture that had been missing entirely. The top of the funnel focused on demand creation — reaching people who weren't yet actively shopping for furniture but were in the right life stage. The mid-funnel nurtured consideration. The bottom closed with clear, direct offers. This separation was critical: without demand creation at the top, all the retargeting and conversion spend was fighting over a pool of users that was too small to scale.
| Metric | Before | After (Day 90) | Change |
|---|---|---|---|
| ROAS | 1.6× | 2.65× | +65% |
| CPA | 100 | 62 | -38% |
| Monthly Revenue | 1× | 2.3× | +130% |
| Video/UGC in creative mix | ~10% | ~65% | Rebuilt |
| Budget scaling | Stopped at 2× | Stable up to 4× | Scalable |
The results came progressively — not all at once. The structural changes in the account took effect within the first 2–3 weeks, stabilizing performance and reducing wasted spend. The creative shift started showing impact in week four, as video and lifestyle content began outperforming catalog ads by a significant margin. By the end of month three, the compounding effect of all three changes — structure, creative, and funnel — produced results that exceeded our initial projections.
The highest-performing creatives were room transformation videos — content that showed a space before and after the furniture was placed. These consistently delivered the lowest CPAs and the highest conversion rates across all placements. UGC-style content (real customers showing their purchases in their homes) performed nearly as well, particularly for retargeting.
This project reinforced something we see consistently: in most underperforming Meta accounts, the targeting isn't the problem. The problem is the system around it — fragmented structure that prevents the algorithm from learning, creative that doesn't match the buyer's mindset, and a funnel that tries to convert people who were never warmed up to begin with.
For the furniture category specifically, the most powerful lever was creative. Broad-targeting campaigns paired with the right creative not only reduced costs — they unlocked a scalable structure that the brand can build on long-term. The 90-day window was the beginning, not the ceiling.
When we started working with a DTC perfume brand in Germany, the brand already had strong products, an established operation, and active campaigns. Sales were coming in.
But performance was volatile. During certain periods, sales spiked sharply — then dropped just as fast. Two core problems emerged: missed opportunities during high-demand windows, and uncontrolled costs during low season.
| Problem | Root Cause | Impact |
|---|---|---|
| High-season underperformance | Generic campaigns, no event strategy | Missed revenue |
| Low-season cost bleed | Same budget approach year-round | High CPA |
| Wrong creative mindset | Product-focused, not occasion-focused | Low engagement on key dates |
The perfume category is highly event-driven. Valentine's Day, Black Friday, New Year — these moments don't just boost sales, they completely change why the user is buying. The brand was treating all of these as normal campaign periods. We reframed the question:
"Why is the user buying this product right now?"
We concentrated on two key periods with completely different strategies:
| Dimension | 💝 Valentine's Day | 🖤 Black Friday |
|---|---|---|
| User motivation | Emotion, romance, gift-giving | Opportunity, deal, urgency |
| Decision speed | Considered, planned | Fast, impulsive |
| Creative angle | Romantic scenarios, gift unboxing, emotional videos | Aggressive offers, clear CTAs, fast rotation |
| Audience split | Men buying for women (separate creatives) | Broad, high intent |
| Metric | Before | After | Change |
|---|---|---|---|
| Valentine's Day ROAS | Baseline | +72% vs baseline | +72% |
| Black Friday Revenue | 1× | 3.1× | +210% |
| CPA | High (uncontrolled) | -29% reduction | -29% |
| Quarterly Revenue | 1× | 2.4× | +140% |
| Peak performance timing | On campaign day | Pre-event build-up phase | Shifted |
This case shows that in high-seasonality categories like perfume, success doesn't come from ad optimization alone. The real difference came from the right timing, the right message, and the right creative — matched to buyer intent. The same products, the same audience, but a completely different approach produced dramatically different results.
When we started working with this baby stroller brand, they were already active on Meta and had reached a certain sales volume. But growth had stalled. Performance looked stable — yet CPA was above target, and every time the budget was increased, efficiency dropped fast.
The first analysis revealed that the problem wasn't just creative or campaign structure. There was a deeper insight — about who was already in the audience.
| Signal | Finding | Implication |
|---|---|---|
| Conversion source | Majority from warm/existing audiences | Brand trust already built |
| Pixel data | Strong — sufficient signal history | Ready for Advantage+ |
| Creative approach | Product-focused, feature-heavy | Missing trust signals |
| Targeting | Narrow, manual segments | Blocking algorithm from best audiences |
Replaced narrow manual segments with Meta's Advantage+ structure. Let the algorithm leverage strong pixel data to find buyers — both new and returning.
Shifted from product specs to real customer reviews, UGC videos, and mom-in-daily-life content. Answered the buyer's core question: 'Are other moms happy with this?'
A mother buying a stroller isn't just buying the product — she's buying the experiences of other mothers. Review-based and UGC creatives consistently outperformed spec-based ads in both CTR and conversion rate.
| Metric | Before | After (Day 120) | Change |
|---|---|---|---|
| CPA | Above target | -34% reduction | -34% |
| Conversion Rate | Baseline | +41% increase | +41% |
| ROAS | Baseline | +58% increase | +58% |
| Repeat Purchase Rate | Low | Significant increase | ↑ Improved |
| New customer acquisition | Stalled | Sustainable scale via Advantage+ | Unlocked |
In trust-driven categories like baby products, three things matter above all: speaking to the right decision-maker (the mom), letting existing customer data power the algorithm (Advantage+), and building creatives that answer 'are others happy with this?' The combination of strong data, the right creative, and Advantage+ made both new customer acquisition and maximum return from existing audiences possible at the same time.
Our client is a marine tourism operator based on the coast of Oman, offering unique snorkeling and underwater exploration experiences. With a strong emphasis on customer satisfaction, environmental responsibility, and high-quality service, the brand has established itself as a trusted provider of ocean-focused tours.
Their primary objective was clear: strengthen their digital presence, increase direct reservations, and reach potential customers more effectively through paid social advertising. But before any campaigns could launch, there was critical groundwork to be done.
Before the first meeting, we conducted a comprehensive review of the client's digital ecosystem: website user experience, social media presence, customer feedback, and the overall booking and conversion flow. What we found was a brand with real potential but a technical setup that wasn't ready to support paid advertising efficiently.
During our strategic meeting, the conclusion was direct: launching paid ads without first optimizing the website would mean spending budget on traffic that couldn't convert cleanly. The most responsible decision — for the client's budget and long-term results — was to pause on campaigns and build the right foundation first.
Protecting a client's advertising budget is just as important as growing it. Running campaigns on a broken foundation doesn't just waste money — it generates misleading data that makes future decisions harder.
| Area Reviewed | Finding | Action |
|---|---|---|
| Website UX | Booking flow not optimized, slow on mobile | Rebuild before ads |
| Tracking setup | No pixel, no events, no conversion data | Must configure from scratch |
| Payment flow | Stripe redirects off-site — purchases untrackable | Custom solution required |
| Meta Business setup | Incomplete or missing | Build from ground up |
Following our recommendations, the client worked with their own development team to upgrade their WordPress-based website. Our role in this phase wasn't to build — it was to guide. We provided precise technical requirements so that when the development work was done, we could integrate advertising infrastructure without friction.
After approximately 2–3 months of development, the website was ready. And critically, it was ready in exactly the way that would allow our advertising infrastructure to function properly from day one.
With the website ready, we moved into the full Meta advertising setup. This wasn't just 'installing a pixel' — it was building a complete, secure, and properly structured business ecosystem that could support professional performance marketing.
The most technically complex challenge was the Purchase event. The website uses Stripe for payment processing, which redirects users off-site to complete their transaction. Under this setup, Meta's pixel loses sight of the user the moment they leave — meaning every completed booking would go unrecorded.
We solved this with a custom implementation: after payment completion on Stripe, users are automatically redirected back to a dedicated Thank You page on the client's website. The Purchase event is fired on this page — giving Meta a clean, reliable signal for every single booking. No guesswork, no data gaps.
| Event | Trigger Point | Purpose |
|---|---|---|
| ViewContent | Tour detail pages | Interest signal, lookalike building |
| AddToCart | Tour selection step | Intent signal, retargeting seed |
| InitiateCheckout | Checkout page load | High-intent retargeting |
| CompleteRegistration | Account / newsletter signup | Lead generation tracking |
| Lead / Contact | Inquiry form submission | Soft conversion tracking |
| Purchase ✓ | Thank You page (post-Stripe redirect) | Booking confirmation — conversion optimization |
Most brands rush to launch ads before their infrastructure is ready. The result is a cycle of spending without learning — high costs, low visibility, and no reliable data to improve from. This project is a reminder that the most valuable work sometimes happens before the first campaign goes live. A clean, validated, technically sound foundation isn't a cost — it's the single best investment a brand can make before entering paid advertising.
For this marine tourism brand, the foundation is now in place. The next chapter — performance campaigns built on real conversion data — can start with confidence.
This flower shop had everything working offline: a loyal local customer base, strong product quality, and consistent foot traffic. But online? The brand was essentially invisible. The website existed, but there was no real digital growth engine behind it — no structured ad strategy, no performance data to optimize from, and no scalable system to move beyond local sales.
The timing added another layer of complexity. Meta's Androme update had recently limited interest-based targeting — meaning the old playbook of stacking narrow interest audiences no longer worked. The path forward required broad targeting, which put significantly more pressure on creative quality and testing discipline.
| Challenge Area | Situation | Implication |
|---|---|---|
| Online presence | Minimal — offline-first brand | No existing audience data |
| Targeting environment | Post-Androme: interest targeting limited | Broad targeting + strong creative required |
| Scaling method | No prior testing structure | Build ABO→CBO workflow from scratch |
| Conversion funnel | Warm audiences underdeveloped | Build retargeting layers gradually |
The strategy was built on three interconnected pillars: a broad-first, funnel-based audience structure; a disciplined creative testing and scaling workflow; and targeted promotional campaigns aimed specifically at warm audiences.
Broad targeting across Germany + warm audiences built from traffic, cart, and engagement signals + lookalikes from high-intent actions
Similar creatives tested in the same ad set, different styles in separate sets — then winning creatives scaled into high-budget CBO campaigns
Two dedicated promo campaigns launched — targeted exclusively at warm audiences, not cold traffic. Each with custom creatives featuring unique codes
The core of the performance strategy was a structured testing-to-scaling pipeline. Rather than putting all creatives into a single campaign and hoping the algorithm would sort it out, each test was designed to produce clean, actionable data.
This approach meant that advertising budget was never wasted on unproven creatives at scale. Testing costs remained minimal, and scaling happened only after performance was validated. The result: a consistently efficient account that improved month over month.
Four creative formats were tested throughout the campaigns. The results were clear and perhaps counterintuitive for a visually rich product category like flowers.
In the floral category, a single high-quality image of a beautiful bouquet stops the scroll immediately. The product sells itself visually — animation and motion often distract from the core message rather than strengthening it. Simple, clean, product-forward imagery consistently outperformed everything else.
| Metric | Result | Driver |
|---|---|---|
| Online Sales Volume | +35% | ABO→CBO scaling + promo campaigns |
| ROAS | +20–25% above initial target | CBO efficiency + creative quality |
| Add-to-Cart | +30% | Funnel structure + retargeting |
| CTR | +20% | Continuous creative optimization |
| Promo period sales lift | +20–25% | Warm audience promo-code campaigns |
| Best performing creative | Static image — ~30% better | Product-forward simplicity |
In a post-Androme world, creative quality is the targeting. This project proved that broad targeting works — but only when the creative does the heavy lifting. The ABO→CBO workflow removed guesswork from scaling: test small, validate quickly, and only invest big in what's already proven to work.
The most surprising finding? Static images crushed everything else in a visually rich category. Sometimes the simplest, most direct presentation of a beautiful product is more powerful than any production technique — and this project is the data to prove it.
When we started working with this German online auto parts and accessories store, the brand was already running active Meta campaigns and had reached a certain order volume. But performance was below expectations. Traffic existed, product variety was strong, and prices were competitive. Yet conversion rates were low and ad costs were climbing — especially when budget was increased, efficiency dropped rapidly.
The auto parts category works differently from typical e-commerce. Users know what they're looking for. They need a specific part for a specific vehicle. The decision cycle is fast — and intent is high. But the campaigns were working against this.
| Area | Situation Before | Impact |
|---|---|---|
| Creative messaging | Too broad — no vehicle specificity | High-intent traffic not converting |
| Trust signals | Missing — no reviews, guarantee info | Buyers hesitating at decision point |
| Campaign structure | Manual targeting, no Advantage+ | Large catalog underutilized |
| Retargeting | Basic — no DPA for cart abandoners | Revenue left on the table |
In the auto parts category, users don't need inspiration — they need answers. The central question we built the strategy around was: 'What does the user care about most when buying this part?' The answer was clear: getting the right part, confirming it fits their specific vehicle, and trusting the seller.
We rebuilt the entire communication strategy around this insight. Rather than generic product ads, creatives were redesigned to directly eliminate the buyer's uncertainty — making vehicle compatibility, product codes, fast delivery, and trust signals the primary message.
Ads directly highlighted which makes and models the part fits. Eliminated the #1 hesitation point: 'Will this work for my car?'
Customer reviews, guarantee badges, and fast delivery promises added to creatives. Buyers needed to trust the seller — not just the product.
Broad targeting + Advantage+ Shopping Campaigns activated. The large product catalog could now be matched to the right users by the algorithm.
Dynamic Product Ads optimized separately for product viewers and cart abandoners. These two audiences represented the highest-intent users — untouched before.
In rational decision-making categories like auto parts, the goal isn't to convince — it's to eliminate doubt. The user already wants to buy. They just need certainty. Creatives that directly answer 'will this fit my car?' and 'can I trust this seller?' convert dramatically better than any creative that tries to be interesting.
| Metric | Change | Primary Driver |
|---|---|---|
| Conversion Rate | +46% | Trust-first creative + vehicle compatibility messaging |
| CPA | -31% | Advantage+ efficiency + higher conversion rate |
| ROAS | +62% | Combined effect of all changes |
| Order Volume | 2.1× | Advantage+ Shopping + DPA retargeting |
| Top Revenue Sources | DPA + Advantage+ | Drove majority of total revenue in 90 days |
In high-intent, rational categories like auto parts, the advertising job is not to create desire — it's to remove friction. Users arrive ready to buy. The creative just needs to confirm: 'Yes, this is the right part for your car. Yes, you can trust this seller. Yes, it will arrive fast.' Every element of doubt removed is a conversion gained.
On the technical side, Advantage+ Shopping and DPA together created a self-reinforcing system: Advantage+ brought in new buyers efficiently, and DPA recaptured the high-intent users who had already shown interest. The 2.1× order volume growth came not from spending more — but from spending smarter.
This is a premium jewelry and watch brand operating across two European markets: Germany and Switzerland. The product range covers fine jewelry — necklaces, bracelets, rings, earrings — alongside luxury watches. Price points reflect the premium positioning: not fast fashion, not ultra-luxury, but the considered middle ground where quality and craftsmanship meet accessibility.
Both markets speak German. Both buy online. Both respond to similar creative aesthetics. And yet — they are structurally separate markets that require a separate operational approach in Meta advertising.
When a brand sells in multiple countries, the instinct is often to run one campaign targeting all of them. Simple, efficient, less overhead. But for this brand, that approach created a fundamental problem.
Own domain and URL structure. German pricing in Euros. German checkout flow and trust signals.
Separate domain and URL structure. Swiss Franc pricing (CHF). Swiss-specific checkout and shipping.
If a Swiss user clicks an ad and lands on the German URL — they see Euro prices, a German checkout, and potentially wrong shipping options. Conversion collapses. The user isn't confused because the product is wrong. They're confused because the landing experience is wrong. This is a structural error, not a creative one. And it requires a structural solution.
The solution was a clean dual-account structure. One Meta ad account dedicated entirely to Germany. One dedicated entirely to Switzerland. Each account operates independently — its own campaigns, its own budget, its own pixel, its own URL destinations.
The creatives themselves are identical. Both Germany and Switzerland are German-speaking markets, so there is no language barrier, no need to produce two sets of content. A single creative library serves both accounts. The only difference — and it is the critical one — is the destination URL and the account-level geo-targeting.
Swiss user clicks → lands on .de store → sees EUR, German shipping → abandons. German user clicks → routed to .ch → confused. Budget wasted with uncontrolled URL routing.
Swiss user clicks → always lands on .ch store → sees CHF, Swiss shipping → correct experience. German user clicks → always lands on .de store → correct experience. Every click, right landing page.
In multi-market advertising, URL integrity is as important as creative quality. You can have the most beautiful jewelry ad in the world — but if a Swiss customer lands on a German-priced checkout page, you've lost the sale. The dual-account structure eliminates this risk entirely. It is not a workaround. It is the architecturally correct solution for brands operating across separate country storefronts.
Jewelry and watches are inherently visual products. The creative strategy for this brand centered on showcasing the product in the best possible light — literally and figuratively. Clean, elegant photography. Minimal text overlays. The product speaks for itself.
Lifestyle imagery with close-up product detail. Emotional, aspirational framing — jewelry as moments, not objects. Gifting occasions, seasonal moments.
Precision-focused imagery. Product alone, or worn on wrist. Craftsmanship details — mechanisms, materials, heritage. Rational and aspirational balance.
German-language copy runs identically in both markets. Switzerland's German dialect is close enough — no localization needed. One production cycle serves both accounts.
Broad targeting in both accounts. Advantage+ to let the algorithm identify high-intent luxury buyers in each market. Retargeting separately per country.
| Dimension | Single Account | Dual Account (This Brand) |
|---|---|---|
| URL routing | Risk of wrong country landing | Guaranteed correct URL per market |
| Budget control | Shared — hard to split by country | Independent — DE and CH budget set separately |
| Pixel data | Mixed — DE and CH conversions in one pixel | Clean — each country has its own pixel history |
| Performance reporting | Blended — hard to isolate per market | Clean — DE and CH performance tracked independently |
| Creative production | One set | One set — same creatives, both accounts |
The insight from this case is deceptively simple: when you sell in two countries with separate online storefronts, you need two ad accounts — one per country. Not because of language, not because of creative differences, but because of URL integrity. Every user who clicks your ad must land on the correct country's page. This is not optional for brands with country-specific pricing and checkout flows. The dual-account structure makes this a guarantee, not a best-effort.
For this brand, the efficiency gain was also significant: one creative set, one production workflow, two fully independent and correctly routed market presences. That's the structural elegance of the dual-account approach.
This brand operates a food ordering platform serving customers across Germany and Austria. Users can place orders both through the mobile app and directly via the website — the order experience is fully available on both channels. The platform connects hungry customers with a range of restaurants and food options, handling the full ordering and delivery flow.
When we started working with this brand, the question wasn't whether to use Meta ads. They already were. The question was: where should the campaigns be pointed — and what should they be optimizing for?
A brand with both a mobile app and a website faces a choice in Meta advertising: do you run app install campaigns, or do you drive traffic to the website? Both are valid approaches. The right answer depends on where the conversion friction is lowest and where the purchase intent can be captured most efficiently.
For a food ordering platform, the website is not a secondary channel — it's a full ordering environment. A hungry user at 7pm who sees a food ad doesn't want to install an app before they can order. They want to order now. Directing them to the website removes every unnecessary step between the ad and the completed order. The goal is not to grow the app's install base. The goal is to generate orders. The website makes that happen faster.
The campaigns were structured as website traffic campaigns — but with conversion as the north star. Traffic without intent is just cost. Every element of the campaign structure was designed to attract users who were already close to ordering: timing, creative, audience, and landing page alignment.
Campaigns weighted toward lunch (11–13:30) and dinner (17:30–21:00) windows. Weekends treated as elevated-intent throughout the day.
Appetite-triggering visuals. Real food photography, no illustrations. Short, direct messaging: 'Order now', 'Delivered in 30 min'. Urgency without pressure.
Separate campaigns per country. Broad targeting in each, Advantage+ identifies high-frequency orderers. Retargeting for visitors who didn't complete.
One of the most important decisions in this campaign was treating the website as a genuine conversion environment — not just a redirect to the app. Many food delivery brands use their website primarily to push users toward app downloads. We took the opposite approach: the website is where the order happens.
| Element | How It Was Used | Conversion Impact |
|---|---|---|
| Landing page | Ads linked directly to the order page, not homepage | Fewer clicks to order |
| Pixel setup | Purchase event fired on successful order completion | Algorithm optimizes for actual orders |
| User flow | Browse menu → select → checkout → order confirmed | Full conversion path on web, no app required |
| Retargeting | Visitors who browsed but didn't order → retargeted with delivery angle | Second-chance conversions captured |
Food advertising has one job above all else: trigger appetite. The most technically perfect campaign structure fails if the creative doesn't make the viewer want to eat right now. For this brand, creatives were built around a simple principle — show the food, make it real, make it immediate.
Close-up, high-quality food shots showing texture and detail. No generic stock imagery — restaurant-specific content that makes the food feel local and real.
Short messaging that removes thinking: 'Jetzt bestellen', 'In 30 Minuten geliefert'. No long descriptions. Just the hook and the CTA.
Short video clips performed strongly for awareness. Static images with a clear product shot drove the most direct clicks to the order page.
DE and AT campaigns used identical creative assets — German-language copy works for both. Minor copy variations tested for regional resonance.
Food purchase intent has a very short shelf life. Someone who is hungry at 7pm and sees your ad is a hot lead. That same person at 7:45pm — after cooking pasta — is completely cold. The entire ad ecosystem must be optimized for immediacy. Any unnecessary step is a lost order.
Like the jewelry brand in a previous case, this brand operates in two German-speaking markets. The same shared-language advantage applies — one creative set, two markets. However, for a food delivery platform, the local dimension matters more.
For a food ordering platform with both an app and a website, the single most impactful structural decision is: optimize for website orders, not app installs. The website eliminates the install barrier and allows users to go from ad to completed order in the fewest possible steps. When combined with intent-based timing, appetite-triggering creatives, and a properly configured purchase pixel, website traffic campaigns become a direct order generation engine — not just a traffic channel.
The food category is uniquely time-sensitive. Every campaign element must be optimized for the moment the user is hungry. Miss that window, and the conversion is gone. Catch it, and the order is immediate.