What can you say when you can't say everything?
A constraint exercise in UK supplement advertising rules, built around a fictional magnesium supplement.
Read the case study →New AI video tools appear every week. New models, new workflows, new possibilities come with them.
Keeping up with all of that shouldn't become another job for your team.
Less production complexity. More creative possibilities.
What happens when you want to test a new idea?
A new hook? A different spokesperson? A different setting? A different product angle? A completely different creative direction?
Does each variation become its own production project?
Most supplement and wellness teams ship 2–4 ad variants a month: enough to keep the lights on, not enough to stay ahead of fatigue.
Meta's own guidance has consistently rewarded broader creative testing with lower costs and steadier reach. The exact thresholds move as its algorithms change, but the direction hasn't: more tested concepts, not fewer, is where the reward sits.
That gap is where winning ads quietly die waiting for a replacement. It's also why our packages are built for volume well past that baseline: Focused Creative Testing delivers 6 tested assets a month across 3 core concepts; Broader Creative Production delivers 12 across 5. Both already sit well past what most teams manage on their own, regardless of exactly where Meta's bar sits this quarter.
What if changing the idea didn't mean starting from scratch?
Which one would you test first? And which one would you test next?
The 6 formats above are built for someone who's never seen your brand: the job there is to earn a first look.
More possibilities can also mean more production complexity.
How many tools, models and workflows would you need to manage to do all of that yourself?
That's the layer we simplify. We keep track of which tools, models and workflows are worth using so you can stay focused on the result.
We're upfront that this is AI-powered production. That's not a caveat: it's how a small, focused team keeps your account stocked with tested variants every month, without agency overhead.
Real and demonstration case studies, broken down the same way every time: the brief, the constraint, what we tested, what we learned.
A constraint exercise in UK supplement advertising rules, built around a fictional magnesium supplement.
Read the case study →Real client work will appear here as it ships, no invented results in the meantime.
3 hooks · 3 visual directions · 1 recurring character · 3 environments
Make a fictional premium daily magnesium supplement's ad genuinely compelling, without leaning on the obvious sleep, stress or recovery claims.
Supplement advertising isn't a blank canvas. Health and nutrition claims are subject to applicable UK rules, and the product must meet the relevant conditions for authorised claims where those claims are used. So for this exercise, we deliberately avoided building the concept around unsupported promises about sleep, stress, cortisol, recovery or medical outcomes.
Can the product still become compelling creative when the obvious claims are off the table?
The constraint didn't remove the creative possibilities. It changed where we looked for them.
Demonstration exercise only. Final advertising claims must be checked against the specific product, evidence, authorised claims and applicable requirements.
What happens when you have to re-explain your brand every time?
A new freelancer, a new agency, a new tool: each one starts from zero. What's your tone? What claims are off-limits? What hook already got tested and flopped? What worked last quarter that's worth doing again?
Re-briefing that context isn't free. It's a meeting, a doc, a Slack thread: another thing on your plate before a single frame gets made.
On a subscription, that context compounds instead of resetting. We already know what we tried for you, what worked, and what didn't, so this month's production starts from where last month's left off, not from a blank page.
Less re-explaining. More building on what's already been learned.
Which tool handles this shot best this week?
How does the same person show up correctly across every scene?
What happens when something's almost right but not quite?
What do we need to respect before we produce?
Who remembers what you already tried, what worked, and what didn't, so you don't have to re-explain it next month?
What would you test if you had more creative capacity?
The goal isn't simply to produce more content. It's to create more useful creative possibilities to test.
No annual lock-in on either tier, and Focused Creative Testing starts as a single test month before anything longer. No per-avatar billing either: a pre-validated avatar library is included, and a custom avatar built for your brand specifically is a one-time add-on.
A focused set of creative hypotheses to put into the market.
A test month for the service, renewable once at the same £997, before any longer commitment.
A broader creative system for testing more ideas, angles and executions.
Why not just use the same tools yourself?
You can absolutely use these tools yourself.
The question is whether you want to make AI production another operational responsibility, or hand that complexity to a team already tracking it, so you only see the finished, tested options.
I'm Hery Zo. I lead a small, AI-assisted team, mainly focused on UK DTC supplement and wellness brands right now, not a generic AI video tool. Fresh, compliant video, made for how this category actually advertises, delivered every month.Hery Zo, founder
Before Hery Zo, I was a marketing manager running a small team through Laurel Portié's $5-a-day video ad framework: a 90-day cycle testing a handful of long-form and short-form video ads, retargeting whoever watched enough of each to be worth it, and killing what didn't work. The method was sound. Producing enough video to actually run it wasn't: at least three 3-minute ads and three 60-second ads, reshot and retested as results came in, all made by hand because AI video didn't really exist yet. That's the exact bottleneck Hery Zo exists to remove, for short-form video specifically. The three-minute ad is a different production problem, and not one we take on.
That experience also taught me something specific about ad fatigue: a winning ad doesn't stay a winner. The same creative wears out once an audience has seen it enough times, so the real job isn't picking one ad and letting it run. It's having honed variations of what already worked ready before performance drops, not scrambled together after the fact.
We're a small, AI-augmented team, not a generic AI video tool and not a bloated agency. UK DTC supplement and wellness is the category we focus on, so the scripts, formats and constraints we work with are shaped around how this category actually advertises, fatigues, and gets rejected, not adapted from a general-purpose tool. No manufactured urgency either: no countdowns, no fake scarcity. If it's not the right time, it's not the right time.
We also work behind the scenes for marketing agencies. If you already manage supplement and wellness clients but don't want to build AI video production in-house, we produce the same tested concepts under your name: you brief us, we deliver finished, compliant assets ready to hand to your client as your own work. No visible partnership, no attribution, just production capacity you can resell.