← All Case Studies
Performance MarketingMeta AdsRestaurant Industry

Restaurant Group, Lahore.
ROAS: 1.2× to 3.8×.

Two years of underperforming Meta ads. The same budget. A complete campaign rebuild. Three months later — ROAS had tripled and ad revenue had covered the retainer six weeks in.

3.8×
Final ROAS (from 1.2×)
90
Days to achieve
42%
CPL reduction
6wk
To cover full retainer from ad revenue
The Challenge

Two Years, Flat ROAS

Tariq Mahmood had been running Meta ads for his restaurant group for two years. The campaigns were live, the budget was funded — roughly PKR 180,000/month — and the ROAS had never climbed above 1.4×. At that level, the ads were covering their own cost and almost nothing more.

He'd tried two previous agencies. Both delivered reports full of impressions and reach. Neither had moved revenue. By the time he reached us, he was close to cutting the budget entirely and writing off paid social as "not for restaurants."

That framing was the first thing we challenged. Meta works for restaurants — but not the way most agencies run it.

The audit

We audited the account before agreeing to take it on. What we found was a campaign structure that had accumulated two years of good intentions and almost no strategic discipline:

01

Seven active campaigns with overlapping audiences competing against each other in the auction.

02

Broad creative — food photography with no offer, no urgency, no call to action.

03

No retargeting structure. Website visitors and past customers were treated identically to cold audiences.

04

Budget allocation set and never revisited — high spend on campaigns with no conversion data.

05

Pixel misconfigured — purchase events were firing but not attributed correctly.

None of this was unusual. Most accounts we audit have at least three of these five problems. The combination, however, meant the algorithm had almost nothing useful to optimise towards.

Approach / Strategy

The Strategy

We proposed a full reset — no "optimising" what was there, because the foundation wasn't worth building on. A new campaign architecture from scratch, built around three audience temperatures:

Cold (Awareness)

Lookalike audiences built from past customers. Creative focused on social proof — real food, real atmosphere, specific dishes people talk about.

Warm (Consideration)

Retargeting visitors who had engaged but not converted. Creative showed offers, time-limited deals, reservation CTAs.

Hot (Conversion)

Past customers and high-intent visitors. Simple, direct creative. Single CTA. No distractions.

We fixed the pixel first — without clean conversion data, nothing else would work. Then rebuilt the campaign structure, set clear budget splits across the three tiers, and created new creative for each tier with specific messaging.

Campaign / Execution

Execution Timeline

Week 1–2

Pixel audit and fix. New campaign structure built. Existing campaigns paused.

Week 3–4

New campaigns launched. Learning phase. Conservative budgets while the algorithm collected data.

Month 2

ROAS reached 2.1×. Reallocated budget away from underperforming ad sets. Began testing creative variants.

Month 3

Retargeting sequences performing well. Hot audience campaigns delivering 5.2× ROAS. Blended ROAS across all campaigns: 3.8×. Ad revenue had covered the full retainer six weeks earlier.

"We'd been running Meta ads for two years and couldn't get past 1.4× ROAS. Pexovia rebuilt the campaign structure from scratch and we hit 3.8× by the end of the third month. Revenue from ads alone covered their retainer within six weeks."

— Tariq Mahmood, Owner, Restaurant Group — Lahore, Pakistan
Results

What Made the Difference

The single biggest lever was the retargeting architecture. The previous campaigns had no warm/hot funnel — every user saw the same awareness creative regardless of where they were in the decision process. Building a proper three-tier funnel with audience-appropriate creative was responsible for most of the ROAS improvement.

The pixel fix mattered too. Meta's algorithm is only as good as the data it's given. Misconfigured pixels create garbage attribution, which leads to garbage budget allocation.

The lesson: Most underperforming Meta accounts aren't failing because Meta doesn't work. They're failing because the structure doesn't give the algorithm what it needs to optimise.

3.8×
Peak ROAS (from 1.2×)
42%
CPL Reduction
90d
Timeline
6wk
Retainer Covered by Ad Revenue
Relevant Capabilities

What actually moved this ROAS.

01

Paid Advertising

Campaign architecture built around audience temperature and clean attribution, not "boosting" whatever's on the page. We fix the data foundation first, structure budget around intent, and let the algorithm optimise against real signal.

  • Meta Ads Manager
See the full service
Your Ad Account

Start with a free
Meta Ads audit.

Running Meta ads that aren't delivering? We'll audit your account free — and tell you honestly what's broken.

Request a Free Audit More Case Studies

More results from our clients.

"Built for businesses that are serious about growth."