How Foodess nearly tripled its output without changing how it looks or sounds
Most case studies open with something broken. This one cannot. The brief was to change nothing and carry more, which is the harder of the two.
Before anyone hires anyone
Four objections this brief had to survive, and what happened here
None of these lines was said to us by this partner, or by anyone. They are the objections this category raises before a contract exists, written in the second person so nothing here can be read as a quotation.
You believe nobody will adapt to your system
The brand's own colour presets were in use from day one, and we balanced light where clarity needed it and nothing else. In place since 17 September 2024.
You expect to lose creative control
The sound hierarchy is the founder's specification, applied verbatim: cooking sounds at 100%, music at 30%. In food content the sizzle is the argument, so it is not ours to reweight. In place since December 2024.
You have found agencies cold and freelancers chaotic
Raw footage to finished web video in under a day, documented, delivered every month footage arrived. The working norm is two to three days per recipe. Since October 2024.
You think the category is too saturated to grow
Four of five peer recipe accounts, the largest around 2.7 times Foodess's size, are adding between 0.00% and 0.07% followers a month, so the objection is largely right: the category is flat. Foodess grew anyway. The rate is not printed here, because the September 2024 start value is the partner's to confirm and they have not confirmed it. It sits in the table below, marked pending. Public peer data, captured 25 July 2026.
The brief
A working system, not a broken one
There was nothing to rescue
The brief, in effect: do not fix anything. Carry more. Keep the look, keep the sound, keep the rhythm, and help a brand in a saturated category publish more, so Instagram keeps sending readers to the website.
A rescue forgives change. A working system does not, and every adjustment you make to one is a risk you introduced yourself.
The goal underneath the request
The request was post-production. The goal, fixed in the first conversations, was traffic from Instagram to the website. Not more views: viewers who land on the recipe.
Formats built for the platforms that actually refer readers. Delivery timed so a seasonal recipe never misses its season. Packaging that respects the appetite-first grammar of food content, which is a real constraint and not a style preference.
The system
What we docked into, and what we added
Brand notes first, then work
The partner's own colour presets from day one, light balanced only where clarity needed it, and no house style applied over the top of a house style that already worked.
One shoot, five assets
Every recipe becomes four video formats: a Reel of around twenty seconds, a full-length vertical, a horizontal web video and a text-free web version. Plus a blog photo set graded in the brand's own preset. The website and Instagram are fed from a single shoot.
Delivery at the speed of seasons
Raw footage arrives monthly in one batch. Finished recipes leave one by one, as each is done, which is the partner's preference and the reason a July recipe publishes in July rather than in the batch that clears in August.
Friction becomes system
A delivery once included a format the brand does not use, and the filenames made finished videos hard to find. The fix did not stay a favour. The naming convention changed to Recipe Name Web, Vert, Reel, and went into the standard operating procedure the same week.
The invisible layer
When the brand's YouTube channel came back to life we shipped music licence certificates before anyone asked, and a licence file now travels with every video folder. The partner has never had to argue with a copyright claim.
The arithmetic of a shoot day
What one shoot is worth, and what most brands publish from it
A shoot day costs the same whether it produces one asset or five. Set what you film, then set what you publish from it.
The counts are the pattern we run on this account. The waste column shows what happens when a shoot day produces one asset, which is the default almost everywhere.
Where it strained
Two tests we did not choose
The partner's own production paused twice
There was no footage for months at a stretch in 2025. Nothing changed at our end, and when shooting resumed nothing needed renegotiating.
A new pair of hands in an established aesthetic
In January 2026 the founder hired a videographer, precisely so more recipes could flow. A new person shooting inside a settled look is where most systems wobble.
Here it produced the opposite. We complimented the new videographer's style-matching on a video the founder had shot herself. The output had stopped revealing its source. Within months the videographer was coordinating with us directly.
No same-period growth data exists publicly for comparable accounts, so we have not invented any. What the public record shows is above, with its capture date attached.
The 2.5-million-view Reel matters less than it looks. It came out of the same pipeline as the dozens that did not go viral. Repeatable systems produce outliers. They just do not depend on them.
Where it went next
Scope that grew the quiet way
The partnership moved to a monthly subscription in January 2026, and our team doubled on the account in June. When YouTube reactivated we prepared the licensing groundwork and laid out channel-management options. For now the founder's own assistant covers scheduling; when the channel is ready for more, so are we.
Two fellow food bloggers came to BUBI on the founder's recommendation.
The audience belongs to the recipes. Foodess's food science and carefully engineered overlays are the engine of it; our system is what let that engine run at close to three times the output without losing its character.
Every figure on this page, sourced
Start with a free audit
We look at what you have, tell you what we would change, and agree the expected result before anything begins.
Start a free audit