--- title: "A user costs $30 and returns $50. Yuliya Lennox says that’s bad." description: "Yuliya Lennox spent a decade scaling apps—and learned that a campaign making $20 per acquired user can be the clearest sign its creative is not good enough." language: "en" publishedAt: "2026-07-22T13:07:46.931Z" updatedAt: "2026-07-22T13:07:46.931Z" authors: - name: "David Barnard" url: "https://www.revenuecat.com/blog/author/david-barnard" category: "Growth" categoryUrl: "https://www.revenuecat.com/blog/growth" canonical: "https://www.revenuecat.com/blog/growth/yuliya-lennox-solid-starts-sub-club-podcast-2026" --- # A user costs $30 and returns $50. Yuliya Lennox says that’s bad. Yuliya Lennox spent a decade scaling apps—and learned that a campaign making $20 per acquired user can be the clearest sign its creative is not good enough. ## Table of contents - [A $20 margin can hide mediocre creative](#a-20-margin-can-hide-mediocre-creative) - [BetterMe’s ugliest ad was the one David remembered](#bettermes-ugliest-ad-was-the-one-david-remembered) - [The No. 2 language app kept its founder in every marketing meeting](#the-no-2-language-app-kept-its-founder-in-every-marketing-meeting) - [Sell the idea before you vibe code the app](#sell-the-idea-before-you-vibe-code-the-app) - [A $10 test market can save a $1,000 US experiment](#a-10-test-market-can-save-a-1000-us-experiment) - [Guest links](#guest-links) ## **A $20 margin can hide mediocre creative** When an app starts spending its first $100 or $500 a day on performance marketing, a stable return feels like success. A user costs $30 to acquire and returns $50. The campaign is profitable. Why not scale it? Yuliya argues that this is exactly when a team should become suspicious. “This is bad for you actually because you stop trying and you think, oh, my LTV ROI, whatever, it looks great. I'm doing everything fine. You are not doing fine. You are doing very, very badly.” [Video](https://www.youtube.com/watch?v=jzlPf100vy4) A predictable margin can convince a team that it has solved acquisition before it has found a genuinely exceptional creative. The stronger signal is not steady performance across several ads. It is the outlier that changes the curve: CPA falls sharply, or purchases triple in a day. That is the difference between an ad that backs out and an idea that can move the business. Early performance marketing should search for those step changes, not settle into comfortable averages. ## **BetterMe’s ugliest ad was the one David remembered** David Barnard still remembers a BetterMe ad he saw roughly a decade ago. It featured a deliberately unpleasant illustration of belly fat hanging over the front of someone’s body. It was not aspirational fitness advertising. It was visually jarring—and that is why it stayed in his memory. Yuliya calls it “the best ad ever.” The polished images of slim people exercising disappeared into the category. The ugly illustration broke the pattern of the feed. “They want something that catches them on the first glance.” — Yuliya Lennox For an app selling a function, the first job of an ad is not to look like a campaign from an established consumer brand. It is to make the product and the pain it solves obvious in the first second. Brand polish is valuable only after the viewer understands what is being sold. There is still a limit. At Solid Starts, a baby-feeding brand built around trusted expert advice, the team tested a creative featuring babies made out of fruit. The image attracted attention, but it also prompted parents to question whether they should trust the company feeding their children. The lesson is not that every app should make every ad ugly. It is that most teams invoke “the brand” too early. Performance creative should have to prove that it damages trust before polish becomes the default constraint. ## **The No. 2 language app kept its founder in every marketing meeting** Founders often look for a senior marketer, agency, or CMO who can take growth off their plate. Yuliya’s experience suggests that delegation works better after the founder has developed their own feel for the market. She points to Eva Learn English, which reached the No. 2 position behind Duolingo. Its founder attended every marketing meeting, developed creative concepts, and stayed close to how the product was communicated. “It's not hard at all to understand marketing, but to feel it with your own hands, this is something that every founder needs to know.” — Yuliya Lennox A founder does not need to remain the company’s permanent media buyer. But they do need enough firsthand experience to recognize a strong message, challenge a weak assumption, and know whether the team understands why customers buy. An agency can execute the work; it cannot manufacture that intuition on the founder’s behalf. ## **Sell the idea before you vibe code the app** Cheaper development has made it easier to build a complete product before proving anyone wants it. Yuliya recommends reversing that order. “Just start with marketing, sell a PDF, sell an idea, sell something and then refund because you don't have a product.” — Yuliya Lennox The tactic is intentionally uncomfortable. Put the promise in front of a real audience, ask people to pay, and learn whether demand exists before investing months in development. If the product does not exist yet, refund the buyers. The point is not to collect revenue; it is to replace encouragement from friends and family with evidence from the market. This also forces a founder to define the business they are trying to build. A niche app might be perfectly capable of supporting a $20,000-a-month lifestyle business while remaining unsuitable for venture-scale economics. Marketing the idea early exposes that distinction before the cost base is locked in. ## **A $10 test market can save a $1,000 US experiment** Localization is usually treated as a revenue-expansion project. Yuliya also sees it as a way to make experimentation cheaper. AI can translate a service in hours, making it practical to test more countries without a long localization project. Teams can use markets with lower CPMs and similar conversion behavior to learn which creative concepts deserve a more expensive US launch. “You need to test 500 ads per week. And if you do that on US CPMs, you will definitely go broke.” — Yuliya Lennox Yuliya says a comparable market such as Indonesia can sometimes turn a 1,000−a−day US test into a 10-a-day experiment, giving a team directional evidence before it pays US CPMs. Cheap translation is not the same as local understanding. Replika made a serious push into Japan and even had a Japanese speaker on the team, but the localization still failed to gain traction. The practical play is to use international markets to widen the testing surface while recognizing that language alone cannot reproduce cultural context. In [the full episode](https://www.youtube.com/watch?v=jzlPf100vy4), Yuliya and David also discuss how organic growth can become a trap, why subscription optimization can slide into black-hat behavior, and why running three apps at once prevents the war-room focus required to scale. ## **Guest links** - [Yuliya Lennox on LinkedIn](https://www.linkedin.com/in/yulichkus)