Five brands. Five places growth was stuck.

What the numbers said · Kicks Machine

A marketplace for luxury sneakers and designer bags, on a margin of only 12–18%. Meta’s ROAS was 3.5x — against the 5x it needed to break even. Every sale lost money.

For months this was read as a traffic problem. It wasn't. People were arriving. They just weren't buying.

Also: No ad testing, one message for every buyer, the homepage used as the landing page, and a huge catalogue barely advertised.

Where it broke

The website didn't deliver what the ad promised.

Heavy discounts on luxury put one question in the buyer's head — "are these real?" Nothing answered it, so they left.

What we changed

1. 01Sell trust firstAds in the order a buyer meets the brand: who we are, a real shop you can walk into, then the reason to buy today.
2. 02A landing page that answers the questionAuthenticity proof at the top, EMI, a mini-cart. The homepage stopped doing a landing page's job.
3. 03A tidier accountFive campaigns instead of a scattered set-up. A catalogue campaign, switched on for the first time. Sneakers split from bags.

The ads themselves · tap any to watch

Watch the ad ↗Stage 1 · never heard of you“Discover the world of Kicks Machine”Global brands under one roof. Earns the benefit of the doubt first.Watch the ad ↗Stage 2 · is it real?“Find a store near you”A real shop, real stock, real people. The most direct answer there is.Watch the ad ↗Stage 3 · ready to buy“Year End Super Steal — up to 55% off”Limited pieces, a deadline, a reason to act today.

The result

From losing money on every sale to 10X at peak.

ROAS · scale 0–6x

Profitable in two months.

Return on ad spend

Before	After	Break-even
3.5x	5.3x	5x

scale ₹0–5,000 · lower is better

Down 35%.

Cost to win one customer

Before	After
₹4,300	₹2,800

7X+Catalogue campaign — never tried before

10XPeak during the sale

5→7XGoogle, then spent more without losing it

The month it all landed · Shopify, storewide

Swipe to read the screenshot →

₹2.48CrGross sales, Nov 2025 (storewide, not Meta-attributed)

1,719Orders

21.97%Returning customers

What the numbers said · Wanderlooms

Riding gear and travel wear, made in India. Not a rescue job — a good brand at its own ceiling. Profitable, with ten paise of room. The budget had sat frozen at ₹16,000 a day for months.

Nothing was broken enough to fix. That's the account that stays stuck.

Also: Budget split so thin no campaign ever learned, and order size left to drift.

Where it broke

Nothing created enough want to act.

The ads were seen and watched. There was just no reason to buy today rather than in March — no offers, no deadlines, no seasonal plan.

We took over in September and spent two months getting it winter-ready.

What we changed

1. 01A reason to buy nowA winter range, an offer with a deadline, and a calendar that gave every month its moment.
2. 02A bigger basketThe windcheater led the range and pulled the average order up to ₹2,200.
3. 03Shot with the founderRiders trust riders more than marketers.

The ads themselves · tap any to watch

Watch the ad ↗The offer“Block the wind, not your style”20% off the Odyssey windcheater — the deadline that never existed.Watch the ad ↗The product“All-weather riding essential”Water resistant, warm, built for rides — the reasons to pay ₹2,200.

The result

Doubled the spend and raised the return.

What the numbers said · R·O·a

Stuck at three sales a week. The five gates came back clean — the ads were doing their job. The problem was where the brand was advertising.

Brands bid against each other for every slot on Instagram. In national skincare that bidding got so expensive one visitor cost more than the brand could make back — hence 20–30 visitors a day.

The ladder read

Stop bidding where everyone else is bidding.

Any agency whose answer is always "make more ads" is selling you something.

Every national brand was chasing the same English-speaking customer. Malayalam and Tamil, with before-and-after proof, were wide open.

What we changed

1. 01Malayalam, not EnglishA real customer, his own language, his own results.
2. 02After the clickClinical results on the product pages, WhatsApp follow-ups to bring them back.

The ads themselves · tap any to watch

Watch the ad ↗The unlockMalayalam, not EnglishA real customer, his own language, his own results.

The result

Same ads. A market nobody else was bidding for.

What the numbers said · Innerwear & activewear

A premium label that grew on the back of its Shark Tank moment. By the time we came in, ads aimed at people who already knew the brand were paying back less than they cost, and reaching anyone new was very expensive.

Almost all the money sat in one catalogue campaign, mostly re-selling to people who had already visited. And every sale came from Meta — if Meta wobbled, so did the business.

Where it broke

People had stopped stopping.

The Shark Tank ads had done their job and then some. Shown for months, they'd become wallpaper. New people never got past the first second.

What we changed

1. 01New openings, new peopleFresh angles and hooks, each built for a different kind of buyer, instead of one story told to everyone.
2. 02Offers with a pullBuy 2 Get 1 and Buy 1 Get 1, so there was a reason to act today.
3. 03Not just MetaA small Google campaign on the brand's own name — competitors were bidding on it — and Google Shopping for people already searching.
4. 04Proof moved to the pageThe Shark Tank badge went from the ads to the landing page, where it reassures buyers instead of tiring them. A “pairs well with” row lifted order size.

The result

New openings, and the account started growing again.

What the numbers said · Ritwika's

The return had sat at ₹2 for three months. The account was split into so many "testing" campaigns that none of them ever settled.

Every time an ad started winning, it was moved to another campaign — which reset it and threw away what it had learned.

Where it broke

They clicked, looked at the price, and left.

Fewer than 1 in 10 visitors added anything to cart. The pieces cost more than other decor brands — rightly, they're handmade — but the site didn't show why. No reviews, no ratings, no one using the product.

What we changed

1. 01Fewer, steadier campaignsOne for new people, one automated, one for people who already knew the brand. The redundant “testing” campaigns went.
2. 02Leave the winners aloneWinning ads stayed where they were, so the campaigns could settle and the numbers meant something.
3. 03Proof on the pageReviews, ratings, and videos from creators and customers — the reasons to pay a premium, right where the price is.
4. 04New buyers, through new adsGrowth came from ads built for new kinds of buyer, not from endless settings changes.

The result

A steadier account, and a better return.