Food & beverage case study: ₹20L to ₹34.7L monthly revenue in 6 months

New ads per month

1. 1

   Research & offer · Month 1

   What we do

   We open with set-up work on the established food and beverage store, since the brand in this scenario named no single problem. We open with three documents: a website audit, a creative brief and a media schedule by month. We build the account in layers so each can be read on its own: testing, scaling, new-buyer prospecting and cart retargeting.

   Why

   An established food and beverage brand grows monthly revenue in 6 months, from ₹20L to ₹50L (2.5x). That is faster than nine in ten of our measured accounts grew over the same time, so the case study below is on what that fastest tenth reached and shows where it lands against the number this scenario calls for. The brand in this scenario named no single problem, so the case study starts from the numbers. Later budget calls need something written to be judged against. Layers keep each budget line readable, so a weak campaign cannot hide inside a strong one.

   How it works

   The brief is agreed first; spend follows it. Retargeting sits next to prospecting and never replaces it.

2. 2

   Measurement & reviews · Month 1 to 6

   What we do

   We track ad-platform revenue beside store revenue and net sales after returns, cancellations and tax. We agree a written number for every month on the way from ₹20L to ₹50L, and start each review with the month-to-date figure against it.

   Why

   No starting return on ad spend was given; the starting return is what measured food and beverage stores of that size hold. Returns, cancellations and tax cut logged revenue, so the numbers are set on net sales. A shortfall is caught in the month it happens, not at the end.

   How it works

   Net sales, not platform revenue, decide each budget step. Each budget step is argued against the written number.

3. 3

   Creative testing · Month 1

   What we do

   We test with video, catalogue ads and UGC and creator video first, rising to several dozen new ads a month by the final month, with carousel watched closely since it returns less than the rest. We run one broad video campaign of the core products, with a customer testimonial, and give it most of the budget. The learning month runs at a low daily budget and moves up only when orders come through. We give each lead product its own campaign and read it weekly.

   Why

   An everyday product sells best when the algorithm is free to find its buyers. The first rupees are for finding the buyer, not for volume. A product nobody buys is visible within a week when it has its own campaign.

   How it works

   A video that takes off gets the budget; tests that do not convert are cut. Only confirmed orders at the low budget unlock the next step. A product that does not sell is paused inside the week. The number of new ads grows with spend; video makes up the largest part and catalogue ads the next.

4. 4

   Scaling · Month 2 to 4

   What we do

   Through Month 2 to Month 4, we push as far toward ₹50L as return allows: the budget rises gently and return dips, and Instagram Reels carries the most spend and Facebook Reels the next. We raise budget only while return on spend holds, and cut it when return drops. We scale into the festive and wedding calendar with seasonal collection campaigns.

   Why

   Across Month 2 to Month 4, the budget rises gently, and return on spend dips. In measured accounts, scaling spend raised cost per order and lowered return on spend, so each step waits for return to hold. Demand is seasonal and regional, so spend moves with the calendar.

   How it works

   Each month's budget is set by the return the last one earned. Spend shifts into each season and between regions with the calendar. Instagram Reels carries the largest share of spend in the final month, with Facebook Reels next. Most spend reaches people who have not bought yet; past visitors return more per rupee.

5. 5

   Steady state · Month 5 to 6

   What we do

   From Month 5, we hold the gains and push toward ₹50L only as far as return allows. We refresh tired ads by mixing old and new creatives, and keep a creative bank. We hold the numbers on net sales as volume grows, since returns rise with it.

   Why

   Growth slows from Month 5, still short of ₹50L. Spend still grows, at a slower pace than during scaling. In one account we measured, click-through fell before revenue did; refresh is how that is avoided. Net sales ran below logged store revenue in our measured accounts.

   How it works

   New creatives mix with proven ones rather than replacing them all at once. Each review opens with net sales against the month's number.

1. Month 1

   First, the set-up: a small daily budget carries the first ads, and store orders are matched to tracking. The website audit, creative brief and media schedule are shared. Net sales after returns are matched to logged revenue.

   • Revenue ₹19.8L
   • ROAS 3.76x
   • Ad spend ₹5.3L
2. Month 2

   Scaling starts: each budget move is read against the return it bought. Spend holds, and return holds.

   • Revenue ₹20.9L
   • ROAS 3.66x
   • Ad spend ₹5.7L
3. Month 3

   The seasonal collection campaign leads this period's spend mix. Spend steps up, and return holds.

   • Revenue ₹25.7L
   • ROAS 3.54x
   • Ad spend ₹7.3L
4. Month 4

   Budget shifts to the products that sold this period.

   • Revenue ₹29.7L
   • ROAS 3.52x
   • Ad spend ₹8.4L
5. Month 5

   The case study moves into its steady phase, leaning on refreshed ads and repeat buyers.

   • Revenue ₹33.9L
   • ROAS 3.26x
   • Ad spend ₹10.4L
6. Month 6

   The creative bank supplies this period's refresh. Spend holds, and return holds. Revenue ends below ₹50L, the number this scenario calls for, because the case study follows what the fastest tenth of our measured accounts reached. Each order costs more by the end than it did while learning.

   • Revenue ₹34.7L
   • ROAS 3.33x
   • Ad spend ₹10.4L

1. 1

   Refreshed tired ads by mixing old and new creatives; one product ad returned to 24.8x

2. 2

   Monthly review calls that start with what has not worked

3. 3

   Tested UGC for one product and a UGC lookalike audience, then cut both