How Lumora Skin Hit $410k/mo at a 6.1x Blended ROAS

So when Lumora Skin first came to us they looked, on paper, like a healthy little DTC brand, right, they were doing around $90,000 a month, the product genuinely worked (a three-step barrier-repair routine for sensitive and reactive skin), the reviews were real and glowing, and the founder, a former esthetician, actually knew more about the skin barrier than anyone I had talked to in the category, and yet the business was quietly bleeding margin, and let me be very honest, the reason was painfully common.

Basically the entire machine ran on paid acquisition, right, something like 88% of revenue was coming straight off Meta and a little bit of Google, and the catch here is that when you are 88% dependent on the ad account, you do not own a business, you rent one from the algorithm, and the rent goes up every single quarter. Their blended ROAS had slid to 1.9:1 over the prior six months, their blended CAC had crept up to $58 against an average order value of $74, and their first-order contribution margin was basically zero by the time you paid for the product, the shipping, the packaging, and the ad that brought the customer in. They were growing top-line and going backwards on profit at the same time, which is the worst trap in DTC.

The content side, and this is the part that really mattered to me, was a graveyard. They were posting maybe four or five times a week across Instagram and TikTok, it was mostly product-on-a-marble-counter flatlays, the occasional repurposed influencer clip, a few stock-feeling "5 tips for glowy skin" carousels, and the result was a following of about 11,400 that did not move product and did not really know the founder existed. Their organic reach was sitting around 140,000 impressions a month, their content was not native to any single platform (the same square video got dumped on Reels, TikTok, and Shorts with the watermark still on it), and crucially the founder, who was their single biggest asset, the actual differentiator, the only reason a skeptical customer would trust an unknown skincare brand over a Sephora shelf, was nowhere in the content.

The way I see it, that was the whole problem in one sentence, right, the most trustworthy thing about the brand was invisible, and so every single sale had to be bought, because nothing was being earned. And it was costing them in three ways that compounded on each other. First, obviously, the rising CAC was eating margin, they were spending roughly $52,000 a month on ads to pull about $90,000 in revenue, do that math and it is not a business, it is a treadmill. Second, the lack of organic trust meant their conversion rate from cold traffic was stuck at 1.4%, because a stranger landing on a skincare PDP from an ad has no reason to believe you, and you cannot retarget your way out of a trust problem. And third, and this is the silent killer, their repeat purchase rate was only 19% inside 90 days, which for a consumable skincare routine is genuinely terrible, and it was low precisely because there was no ongoing relationship, no reason to stay, no founder voice in their inbox or their feed reminding them why they bought in the first place. So they had a leaky bucket on the front end and the back end at the same time, and they were trying to fix it by pouring more paid water in the top, and that, at the end of the day, is exactly the kind of situation our flywheel is built to fix.

Okay so here is what we built, and I want to be specific because the specifics are the whole thing, right, anybody can say "post more content", that is useless advice, what actually moved the numbers was the system.

The core of it is the content flywheel, and for Lumora the input was one focused recording session a month with the founder, that is it, one day, and let me be very honest, getting a busy founder to commit to one real shoot day a month is easier than the "post every day yourself" guilt-trip that every agency hands them and that nobody ever sustains. We did a half-day studio block and a half-day at-home-bathroom-counter block, because skincare lives in two registers, the credible-expert register and the this-is-my-actual-routine register, and you need both. Out of that single day we pulled raw material for 32 to 38 platform-native assets a month, and I mean native, not the same clip cross-posted, right, vertical Reels and TikToks and Shorts cut and captioned and hooked differently for each platform, plus LinkedIn founder posts (yes, LinkedIn for skincare, I will explain), plus long-form YouTube, plus static carousels and email-ready stills.

The asset mix was deliberate and it was built around the actual buying objections in sensitive-skin skincare, because the catch here is that in this category the customer is not lazy, she is scared, she has been burned by products that wrecked her barrier, and so the content has to do the de-risking before she ever sees a price. So the mix was roughly: 40% education (the founder explaining the skin barrier, what "fragrance-free" actually means, why purging is not breaking out, ingredient teardowns), 25% proof (before-and-afters with real timelines, customer-story reaction videos, the founder reading a one-star review and explaining honestly who the product is not for, which builds more trust than any five-star ever could), 20% founder-story and POV (why she left her clinic to build this, the formulation failures, the boring honest operator stuff), and 15% direct product and offer content. That ratio is the thing, right, most brands run it inverted, 80% product, and then they wonder why nobody cares.

On hooks and packaging, this is where the platform-native part earns its money, because the same idea needs a totally different opening on each surface. For Reels and TikTok we led with pattern-interrupt and fear-relief hooks ("stop using this if your skin stings", "the barrier mistake 90% of you are making"), for Shorts we leaned into the search-intent long-tail because Shorts gets fed by YouTube's engine and people literally search "how to repair skin barrier", and for LinkedIn we ran the founder as an operator and a formulator, which put the brand in front of other founders, press, and retail buyers, and that single channel is how the wholesale conversations started later. We hard-rule no watermarks, captions burned in for the 85% who watch on mute, and a hook tested in the first 1.2 seconds or it gets recut.

Distribution is the part people skip and it is the part that compounds, right, we did not just "post", we distributed everywhere the content compounds and then we fed the winners back into paid. So every asset went out organically across Reels, TikTok, Shorts, LinkedIn, and YouTube on a native cadence, and then, and this is the operator move, the organic winners (anything that beat a 3-second-view-rate and save-rate threshold) got handed to their media buyer as ready-made paid creative, because UGC-style founder content that already proved itself in the wild is the single cheapest way to drop a Meta CAC. We also built an email and SMS layer off the same shoot, so the trust-building did not stop at the feed, it followed the buyer into the inbox, which is what actually fixed the repeat-purchase leak.

And the whole thing ran on a simple monthly loop: one shoot, then 30-plus assets cut and scheduled, then distribute native everywhere, then read the data at the end of the month, then the next shoot brief is written by the data, not by vibes. That is the flywheel, right, content does the trust-building before the sales conversation, so by the time someone hits the PDP or the founder's DMs or an abandoned-cart email, they already feel like they know her, and warm traffic converts at a multiple of cold, every time. Basically we made the most trustworthy thing about the brand the most visible thing about the brand, and then we made it work for the ad account instead of against it.

So let me give you the actual numbers, because this is where it gets fun and this is the part that matters, right, vibes do not pay for inventory, results do.

Over the eight months Lumora went from $90,000 a month to $410,000 a month, that is a 4.6x on monthly revenue, and the total revenue across the engagement came in at roughly $1.82M against a baseline run-rate that would have done maybe $720k if nothing had changed, so call it about $1.1M in incremental revenue directly attributable to the engine. The investment in Pixel Samy across the eight months was $148,000 all-in (retainer plus shoot production plus the paid-creative handoff support), and so on incremental gross revenue alone that is a 12.3x ROI, and honestly the way I actually think about it is on margin, because we did not just grow the top line, we fixed the unit economics underneath it, which is the whole point.

Here is the unit-economics story, because this is the one that should make every DTC founder lean in. Blended ROAS went from 1.9:1 to 6.1:1, right, that is not a tweak, that is a different business. Blended CAC dropped from $58 to $33, a 43% cut, and that happened for two compounding reasons: one, the founder-led organic content was acquiring customers for basically the cost of a shoot day, and two, the organic winners we fed into the paid account dropped the cost on the paid side too, because authentic founder UGC that already proved itself organically is the cheapest creative a media buyer will ever touch. Cold-traffic conversion rate went from 1.4% to 3.6%, which is a 2.6x lift, and that is the trust effect in one number, right, when someone has watched the founder explain the skin barrier honestly three times before they ever see the ad, they do not need convincing, they need a checkout button.

And then the back end, which is where the real money in DTC actually lives. The 90-day repeat-purchase rate more than doubled from 19% to 41%, and because the content and the email and SMS layer were all built off the same shoots, the relationship kept going after the first order, so LTV climbed from roughly $112 to about $214 across the engagement, basically a 91% increase. When your LTV nearly doubles and your CAC drops 43% at the same time, your LTV-to-CAC ratio goes from a scary 1.9 to a genuinely healthy 6.5, and that, at the end of the day, is the number that tells you whether you have a business or a treadmill.

The channel breakdown by Month 8 is worth stating plainly. Paid dropped from 88% of revenue to about 47%, organic and owned (email, SMS, organic social, YouTube search) carried the other 53%, which means the brand stopped being a hostage to the ad account. Monthly organic reach went from 140,000 to 9.8 million in the final month, total reach across the eight months was 31.4 million, and the top three founder videos alone did 2.1M, 1.6M, and 1.4M views and are still pulling traffic today because they live on Shorts and YouTube where evergreen content keeps getting served. Total watch time across YouTube and Shorts crossed 410,000 hours, and that is 410,000 hours of a stranger sitting with the founder building trust, for free, forever.

There is also the stuff that does not show up cleanly in a ROAS report but is arguably the most valuable, right. The LinkedIn founder content, which everybody told her was pointless for a skincare brand, generated $2.94M of wholesale and retail pipeline across the eight months and closed two accounts worth a combined $640k in committed first-year orders, which is folded into that $1.82M closed figure. And the brand now owns an evergreen library of 280-plus assets and a proven founder-creative system that will keep acquiring customers in Month 12 and Month 24 long after the campaign spend is gone, and that is the difference between renting attention and owning it.

Let me give you the honest operator perspective on this one, because there were a couple of things that genuinely surprised even me.

The first surprise was how fast the founder content beat the production-heavy stuff, right, we shot some beautiful, polished, color-graded studio content, and it did fine, but the videos that absolutely ran were the ones where she sat on her bathroom counter with no makeup and read a one-star review out loud and said "honestly, this person was right, our cleanser is too gentle if you have very oily skin, here is who it is actually for." That video did 1.6M views and dropped the cold-traffic conversion rate measurably for the next six weeks. The catch here is that in skincare, where everyone is lying with filters and stock before-and-afters, the brand that tells you who it is NOT for is the one you trust, and trust is the entire ballgame in a category built on fear. So the lesson, basically, is do not over-produce, produce honesty.

The second surprise, and this is the one I bang on about, is that the paid account got healthier because of the organic engine, not in spite of the budget shift. A lot of founders think organic and paid are a zero-sum fight over the same dollar, and the way I see it that is exactly backwards, right, the organic content is the R&D lab for the paid account. Every month the buyer got 5 to 8 pieces of creative that had already proven themselves in the wild against a real audience, for free, and so the paid creative testing budget basically went to zero and the win rate on new ads went up, and that is why a 43% CAC drop and a 4.6x revenue jump happened at the same time, which on paper looks impossible and in practice is just what compounding looks like.

The third thing, which the founder said to me near the end and which I think about a lot, is that the engine gave her her time back. Before us she was the bottleneck, posting badly and inconsistently and feeling guilty about it, and one focused shoot day a month turned out to be both more output and less work, because the work was batched and the system did the distribution. That is the quiet unlock, right, founder-led content does not have to mean founder-burnout content.

On what we would do next, the obvious move is to widen the cast, right, eight months in we had proven the founder as the trust anchor, and the next phase is bringing in a second on-camera face (an esthetician on the team) and a customer-creator program so the engine is not single-threaded on one person, and then layering in a TikTok Shop and YouTube long-form review-style content to capture the bottom-of-funnel search intent we were only starting to scratch. We would also push the email and SMS LTV play harder, because at 41% repeat rate there is still a lot of room, and every point of repeat rate is pure margin.

And the reason this compounds, the reason I genuinely believe in the flywheel over the campaign, is that a campaign stops when you stop paying and a library does not, right. Those 280-plus assets and that proven founder system are still serving views and acquiring customers today, the YouTube videos are still ranking for barrier-repair searches, the best Reels are still getting saved and shared, and the brand owns all of it. At the end of the day we did not buy Lumora eight good months, we built them an asset that keeps paying, and that is the difference between a vendor and a partner.

The other thing worth saying plainly, for any founder reading this and doing the math on whether it is worth it, is that the cost curve and the value curve move in opposite directions over time, and that is the whole reason I push the monthly cadence instead of a one-off burst. In Month 1 you are paying for a shoot and getting assets that have not compounded yet, so the ROI looks fine but not magical, right, it was sitting around 1.8x on the first month in isolation. But by Month 6 the same shoot day was producing assets that landed on top of an audience of millions, fed a paid account that was already cheaper, and dropped into email flows that a much larger list was now seeing, so the marginal ROI on Month 6's shoot was north of 9x, and that gap between Month 1 and Month 6 is exactly the compounding I am talking about, the input cost stays flat and the leverage on it keeps climbing. Basically the brand that quits at Month 2 because the first month was "only fine" is the brand that leaves all the real money on the table, and I tell founders that before we ever start, because patience is the price of compounding.

And one last operator note, because it is the thing I am most proud of here, is that we made the founder replaceable in the system without making her invisible in the brand, if that makes sense, right. The trust is anchored in her face and her honesty, but the engine, the briefs, the editing standards, the distribution rules, the winner-to-paid handoff, all of that is now documented and owned by Lumora, so even if we walked away tomorrow they could keep running it, and that, the way I see it, is the only honest definition of a partner, you build the thing so well that they do not strictly need you anymore, and then they keep you anyway because the output is that good.