We research what your customers actually believe — from reviews, Reddit, and their own words — before a single ad gets built. Then we produce it, test it, and tell you exactly why the winner won.
2-minute application. No pitch deck. You'll get a straight answer either way.
Ranked low on general lists. In the skincare accounts we've researched, it's quietly one of the best-performing structures there is.
No client roster yet — we're early, and we'd rather say that than fabricate one. What you can check today: our research, our method, and one honest teardown of your own account.
Real, current video from skincare brands people are actually watching. Click to play — nothing loads until you do, so this page stays fast.
▶
▶
▶
▶
▶
▶Real videos, official brand channels · click to play · not our work, just our homework
Every round of testing feeds the next round of research. Nothing gets produced without a reason attached to it.
Reviews, Reddit, Amazon, your own ad account. What customers actually say, not what we assume.
Personas ranked by how strongly people feel, mapped against every angle and awareness stage you're not reaching.
Concepts built to fill a named gap. Every asset ships tagged: persona, angle, why it exists.
Hook rate first, cheap and fast. Cost per result second, once it's earned the spend. Winners and losers both write back to research.
Not results — we don't have paid case studies yet, and we're not going to invent them. This is the volume and rigor of the research itself.
If a claim on this site can't be traced to something you could verify yourself, it doesn't belong here. That's the standard we're asking you to hold your own creative to.
Right now we build and test video — the format range with the most room to find a genuine gap in most skincare accounts. Not a permanent ceiling, a starting depth.
Full breakdowns of real skincare ad accounts — what's working, what's missing, and what we'd test first. Not client work. Research, done in public, on brands we'd love to work with.
For now, yes, deliberately. It's where the sharpest, most common gap sits across the accounts we've researched, and we'd rather be genuinely good at one range of formats than thin across all of them on day one. Static and founder-content direction are on the roadmap as we take on capacity.
Yes, and it's disclosed — Meta requires it on any AI-assisted ad. What we won't do is generate a synthetic person claiming your product gave them a personal result. That's the one thing audiences reliably punish, and it's a hard line for us regardless of the format.
Reviews — including the bad ones — Reddit threads, Amazon's own buyer-persona data, your ad account's real history. Competitors come last in our process, never first: leading with a competitor scan just means arriving at the idea they already had.
No, and we don't want to. We build the creative and design the test; your buyer executes it. That's also why our guarantee sits on hook rate, not ROAS — hook rate is the thing our work actually controls.
Within 90 days, one concept beats your current best ad on hook rate, or the next month is free. We agree the benchmark from your own account data before any work starts, so there's no argument later about what "current best" meant.
90 days standard. Long enough to test properly. Short enough that we have to earn the renewal every quarter.
You shouldn't, blindly — check the work instead. Everything on this page traces to something public: the Meta Ad Library, our published teardowns, the research standard above. Founding clients get a rate reflecting exactly this stage, in exchange for read-only ad-account access and permission to publish real results once they exist.
CookShip is the parent. Creative is the first thing we're building inside it — skincare ad research and production. Other verticals come later, once this one works.
Send this over and we'll come back with a real teardown of your account — not a sales deck. If it's a fit, we'll book a call. If it's not, you'll get a straight answer and keep the teardown either way.