We love detail, but if you’re short on time, let AI summarize it for you.
Share
CASE SUMMARY
This project in four headings
01
Goals and the Problems We Found
Ad spend was rising while revenue had been falling for three months.
A single account had over a thousand ad sets, with budget sitting at set level rather than campaign level.
Existing customers were not excluded from prospecting and frequency reached twenty on some sets.
02
Our Strategy and Solutions
We reduced the pile of campaigns to one funnel with four layers: cold, warm, hot and customer.
We excluded existing customers and separated overlapping audiences.
We connected the brand’s village production archive to a weekly creative production line.
03
Where We Made the Difference
In four months monthly revenue rose 121% and order volume 92%.
Average basket value rose 15%: customers were stocking a pantry, not buying one item.
Revenue doubled while media efficiency stayed in the same band.
04
How We Did It
We built the structure before touching budget; scale came only once the account could be read.
On Google we cut spend 33%, grew conversions 47% and reduced cost per order 55%.
We opened TikTok as a third channel and broke the dependence on one platform.
Köydengelsin is a food brand that sources honey, butter and ghee from village producers and sells them straight to the consumer. The product was good and the brand was loved, but advertising could not carry the growth. When we took over, spend was rising while revenue had been falling for three months.
Instead of patching individual settings, we rebuilt the whole structure. Four months later monthly revenue was up 121%, order volume 92% and average basket value 15%. More importantly, media efficiency did not drop along the way.
Growing is not the hard part; holding efficiency while you grow is. Doubling the budget to double revenue is no achievement. What we defended was the gap between the two.
Discovery: The Account Had Grown, the Structure Had Not
A single Meta account had accumulated more than a thousand ad sets. Almost the entire budget sat at ad set level rather than campaign level, which made it impossible to read what each audience actually returned. Sets moved on before they could finish the learning phase.
The second problem was the audiences. Existing customers were never excluded from prospecting, so new-customer budget was spent on people who had already bought, and frequency reached twenty on some sets. On Google, brand search was masking the real performance of category campaigns. TikTok had never been opened at all.
The Idea: Structure First, Budget Second
Raising budget on a disorganised account only magnifies the problem. So before touching spend, we built the structure: every stage, from first contact with the brand to repeat purchase, would have its own campaign, its own message and its own measure of success. Only once the structure could be read did we scale the budget.
Execution: We Worked on Five Fronts
We rebuilt the funnel from scratch: the pile of campaigns was reduced to four layers — cold, warm, hot and customer. Retargeting was split into cart abandoners, checkout starters and past buyers, each with its own creative and its own frequency cap.
We cleaned up the audiences: existing customers were excluded from prospecting, overlapping audiences were separated, sets with climbing frequency were switched off, and the crowded set structure was cut down to a number that could actually complete learning.
We turned creative into a production line: the brand’s own village production archive was connected to a weekly creative flow. Every week we tested a new angle, scaled the winner and retired the tired one. In October our best ads returned 50% above the account average.
We restructured Google: brand search was separated from category and retargeting campaigns, which showed for the first time which campaign genuinely brought new customers. We cut Google spend by 33%, grew conversions by 47% and reduced cost per order by 55%.
We opened TikTok as a third channel: so that growth would not depend on a single platform, we carried the Meta funnel logic over to TikTok and connected the catalogue. The channel returned above the account average from its first month and grew its order volume by 82% in the second.
Results: Double in Four Months
Measured against July, the month the work began, revenue rose 82% in August, 120% in September and 121% in October. Order volume rose 92% over the same period. The growth came not from one campaign but from every layer of the funnel working at once.
Monthly revenue change against July 2026
Average basket value rose 15%. That shows the campaigns were bringing in customers stocking their pantry rather than bargain hunters after a single item; the family bundle and the category work fed straight into this number.
Month
Revenue
Orders
Average basket
Media efficiency
July
—
—
—
—
August
+%82
+%67
+%9
−%1
September
+%120
+%94
+%13
+%10
October
+%121
+%92
+%15
−%2
All values show the change against July 2026, the month the work began. Media efficiency expresses the revenue returned per unit of budget spent.
The Mobile Reality and Channel Balance
98.2% of traffic comes from mobile. Every decision, from creative to checkout flow, followed that fact: vertical video, the product on screen within three seconds, a short checkout. With TikTok live, growth is no longer at the mercy of a single platform.
Conclusion
We doubled revenue in four months and held media efficiency in the same band. Efficiency usually falls as budget grows; here it did not, because the growth came from the structure rather than from the spend.
In October revenue held at the September level while spend rose, so at this point of scale efficiency is coming under pressure. Our job in the coming period is to keep growing while bringing efficiency back to the September band: reducing drop-off at checkout, scaling TikTok enough to take load off Meta, and growing repeat purchase through customer lifetime value.
Data covers July–October 2026. At the brand’s request absolute figures are not disclosed; all values are given proportionally.
Let’s talk about where your brand stands today and look together at where it could go. We promise no numbers in the first meeting; we look at your account first and name the problem plainly.