Google Ads
Meta Ads
ROAS Before
2.5×
Losing money on every sale
ROAS After
3+×
Achieved within 2 months
Revenue from Outskirts
60%
Costly, inefficient deliveries
The challenge
Cakes taste great, but are hard to sell on a website.
Think about it – when was the last time you bought a cake from a website? That was the upstream challenge baked into this account before we even touched a single campaign.
"More than 60% of their business was being driven from the outskirts of Mumbai. This made delivery expensive, time-consuming, and often led to angry customers."
Zinque had not been profitable since the beginning. Despite consistent ad spend on both Google and Meta, the unit economics simply didn't work — and no one had identified why.
Zinque had not been profitable since the beginning
Zinque
Patisserie · Mumbai, India
Sells cakes and pastry products exclusively through their website
Daily van deliveries across the city of Mumbai
60%+ of ad-driven revenue came from outskirts — expensive to serve
ROAS sat at 2.5 — below their break-even target of 3×
Had paused Meta Ads multiple times due to near-zero returns
60%
of revenue from outskirts far from their delivery hub, raising costs on every order
2.5×
ROAS — below the 3× minimum needed to cover costs and operate profitably
The Dilemma
A classic catch-22
Fixing one problem would make the other worse. There was no obvious path forward – until we found one.
Option A
Fix the locations
Exclude outskirts and cut costly deliveries. But 60% of revenue disappears overnight. CPCs inside Mumbai city are higher due to heavier competition.
📉 Sales volume collapses
or
Option B
Fix the ROAS
Improve return on ad spend without touching locations. But this means delivery costs remain high, customers stay unhappy, and margins stay thin.
📦 Delivery nightmare continues
We fixed both – the locations and the ROAS at the same time. Here's how.
Strategy – Part 1
Fixing Google Ads
Fixing one problem would make the other worse. There was no obvious path forward – until we found one.
1. Structural campaign overhaul
We audited the existing account and paused all but two campaigns. Rather than patching a broken structure, we built a fresh set of campaigns designed to complement each other – each with a specific role in the funnel.
We also completely overhauled the bidding strategy, aligning it with the algorithm's learning behaviour and Zinque's margin requirements.
Note: We expected Month 1 to be a rough ride. Excluding top-selling outskirt locations while simultaneously discarding 90% of campaigns means administering a shock to Google's algorithm. Short-term pain for long-term gain.
2. Full-funnel friction audit
Poor ROAS rarely has one cause. We mapped every touchpoint a customer hits before purchasing and removed every unnecessary obstacle – from ad copy to the checkout page.
🎯 Targeting
→
📣 Ad Creative
→
🌐 Landing Page
→
🛒 Checkout
As part of this, we designed and built new, campaign-specific landing pages – optimised for conversion, not just aesthetics.
A snippet of the hero section of a landing page we built for driving Google Ads traffic.
In the month of May, and thereafter, we started seeing performance hit the numbers they needed us to hit.
May ROAS
Zinque finally started making money on ads for the first time in over a year, within two months of us taking over the account. What’s more, it was from the locations that were cost-efficient and convenient for them - within Mumbai city.
June ROAS
July ROAS
Strategy – Part 2
Fixing Meta Ads
If Google was a problem, Meta was a bigger one. It had barely driven any revenue for Zinque. They'd stopped advertising on it – multiple times.
We approached Meta with one principle: no creative goes live unless it earns its place. Every piece of content had a reason to exist – a strategic intent behind it.
"We would not release creatives that just looked good. If it wasn't strategic, we wouldn't do it."
As with Google, we turned off every existing campaign – leaving nothing running – and started entirely from scratch. All outskirt locations were excluded from day one.
Month 1
Resetting the baseline
Campaigns rebuilt from zero. Location targeting corrected. Algorithm enters learning phase. Performance is subdued — intentionally.
Month 2
Turning profitable
Algorithm stabilises on the right audience. ROAS crosses the 3× threshold for the first time — from within Mumbai city only.
Month 3 & Beyond
Scaling profitably
ROAS holds above target. Spend begins scaling. Zinque continues working with us to this day.
