Let's Talk
Portfolio of wins

Case Studies

Amazon Ads Quick Commerce Measurable ROAS

01

Amazon Ads

Indian Confectionary Brand

1X → 4XROAS in 3 quarters
Indian Confectionary Brand - Amazon

Scaled ROAS from ~1X to nearly 4X by shifting to high-intent keywords, negatives, and full manual campaign control.

  • Sponsored Products
  • Dynamic bidding ↓
  • Top-of-search focus

When I took over the brand in January 2024, the initial ROAS was just about 1. I focused on high-intent keywords and applied negative keywords to eliminate irrelevant traffic. Stopped running automatic campaigns and completely transitioned to manual campaigns for better control. I prioritized Sponsored Product Ads, used dynamic bidding (down only), adjusted bids for top-of-search placements, and allocated most of the budget to top-performing products.

02

Amazon Ads

Middle Eastern FMCG Brand

14 → 45+Restructure to peak ROAS

Full campaign rebuild: portfolio segmentation, 9 manual structures, and reporting that unlocked Ramadan and May peaks.

25Post-rebuild ROAS
33.65Ramadan ROAS
45+May 2024 ROAS

I inherited ineffective auto campaigns and unstructured manuals, then rebuilt from the ground up — segmenting portfolios, researching keywords, and launching brand / competition / generic campaigns across match types plus product, category, and sponsored brand coverage.

Within three months: Ramadan drove 2.25M revenue at 33.65 ROAS. May 2024 crossed 3M revenue with ROAS over 45.

03

Swiggy Instamart

Candies & Gums

2×ROAS improvement

Exact-match keywords + time-slot pausing turned sub-2 ROAS into a sustained 2x lift by June.

A

Keyword Optimization

Moved all keywords to Exact match for tighter intent and less wasted spend.

B

Time Slot Targeting

Paused low-intent early breakfast hours at the end of May.

By March, Instamart ROAS was stuck below 2 with mixed Broad/Exact match and soft bids. After Exact-only keywords and breakfast-slot pauses, June performance doubled — then the same playbook scaled to two more campaigns.

Key takeaways

  1. Target high-intent time periods
  2. Use Exact match to cut waste
  3. Monitor, experiment, adapt weekly
Scroll to Top