Restaurant Meta Ads Case Study in Pakistan

The Client

A 50-seat casual dining restaurant in Lahore’s Gulberg III area, serving Pakistani fusion cuisine with a focus on lunch and dinner service. The restaurant had been operating for three years in a prime location near M.M. Alam Road, surrounded by corporate offices, retail outlets, and residential apartments. Weekend dinner demand was strong (regular 85-90% occupancy, Friday-Saturday), but weekday lunch traffic was inconsistent and unpredictable—averaging 35-40% occupancy with Tuesdays and Thursdays particularly weak.

The restaurant’s menu included traditional Pakistani dishes with contemporary presentation, plus a range of beverages and desserts. Average check size during weekday lunch was PKR 1,800-2,200 per person, making cover acquisition a critical lever for revenue. The restaurant had an active Instagram page with 8,000+ followers and a Facebook business page, but no structured Meta advertising beyond occasional boosted posts and generic city-wide campaigns that ran primarily in evening hours.

Marketing decisions were made by the owner-operator with limited analytics visibility. Previous Facebook campaigns had produced likes and comments but unclear attribution to actual covers in seats. The restaurant needed a predictable weekday pipeline that didn’t cannibalize already-strong weekend demand and could be measured in terms of real customers, not just engagement metrics.

The Problem

The restaurant’s Meta Ads account had sporadic activity—boosted posts when business was slow, city-wide campaigns with generic food photography, and ad scheduling concentrated in evening hours (6pm-10pm) when competition from other restaurants was at its peak. Campaigns targeted Lahore city-wide with 10km+ radius, reaching audiences unlikely to travel 8km for a weekday lunch. There was no daypart optimization for lunch-decision windows (10am-1pm when office workers plan meals) and no location exclusions for areas too far to convert.

Key blockers:

• Weekday lunch occupancy averaging 35-40% (18 covers/day on a 50-seat capacity) despite prime location
• Monthly Meta spend of PKR 90,000 with ROAS below 1.8x on weekday campaigns
• City-wide targeting waste—audiences 8km+ away consuming budget without converting
• Evening-heavy ad scheduling missing lunch-decision windows (10am-1pm peak planning time)
• Generic food-photography creatives with no offers, urgency, or weekday-specific hooks
• No location-based exclusions preventing spend on areas unlikely to convert (e.g., Cantonment, Township)

The restaurant was struggling with predictable weekday revenue. Weekend dinner was strong, but Tuesday-Thursday lunch was weak despite nearby office density that should have driven steady demand. Previous Meta campaigns produced engagement (likes, comments, shares) but no measurable impact on actual covers. Without proper tracking, the restaurant couldn’t distinguish which campaigns, audiences, or creatives actually drove customers to make reservations or walk in.

Phase 1 — Audience and timing diagnosis (Weeks 1-2)

We began by analyzing the restaurant’s existing Meta performance, customer transaction data, and competitive landscape in Gulberg. The Meta Ads account had basic campaign structure but suffered from geographic waste, poor scheduling, and generic creative. We installed the Meta Pixel and Conversions API to track website visits, menu views, reservation clicks, and estimated walk-in conversions (using time-of-day correlation and reservation data).

First, we analyzed customer transaction data from the POS system to understand current dining patterns. Weekend dinners (Friday-Saturday, 7pm-11pm) were 85-90% occupied with average checks of PKR 2,800 per person. Weekday lunch (Tuesday-Thursday, 12pm-3pm) averaged 35-40% occupancy with checks of PKR 1,900 per person. This revealed a clear opportunity: growing weekday lunch would boost revenue without cannibalizing already-strong weekend demand.

We mapped the restaurant’s primary geography using customer address data and drive-time analysis. 73% of weekday lunch customers came from within 1.5km of the restaurant—primarily nearby office workers and residents. Weekend dinner customers traveled further (median 3.2km) for special occasions. This insight shaped a location-first strategy: target the 1.5km radius heavily for weekday lunch while maintaining broader campaigns for weekend dinner.

Competitive analysis revealed that other restaurants in Gulberg were heavily advertising in evening hours (6pm-10pm) with generic food creatives. Lunch-decision windows (10am-1pm) had less ad competition, creating an opportunity to capture office workers planning meals mid-morning. We identified that Pakistani office-lunch decisions are often made the same morning, with group discussions around 11am-12pm, requiring Meta campaigns to run earlier than typical restaurant ad scheduling.

We tested multiple radius options—1km, 1.5km, 2.5km, and 5km—to find the optimal balance between reach and efficiency. Early testing showed that 1.5km produced the highest ROAS (3.1x) for weekday lunch, while 5km campaigns had lower engagement and higher cost per cover. This validated the location-first hypothesis.

Diagnosis Finding	Insight	Immediate Action
City-wide campaigns wasted 62% of spend on audiences >3km away	Restaurant’s weekday customers are hyperlocal (73% within 1.5km)	Restructure campaigns by radius—1.5km for weekday lunch, broader for weekend
Evening-heavy scheduling (6pm-10pm) missed lunch-decision windows	Pakistani office workers decide lunch between 10am-1pm same-day	Shift weekday campaign scheduling to 9am-2pm
Generic food photography produced likes but not reservations	No urgency trigger, no weekday-specific offer, no clear call-to-action	Test offer creatives (weekday deals, group offers, corporate mentions)
Weekend strong, weekday weak	Opportunity: grow weekday without cannibalizing weekend	Build separate weekday campaign with distinct budget and optimization

Phase 2 — Location-radius campaign build and creative testing (Weeks 3-4)

We rebuilt the Meta Ads account with two core campaign tracks: weekday lunch (Tuesday-Thursday) and weekend dinner (Friday-Saturday). Each campaign used distinct geographies, schedules, and creative approaches. Weekday campaigns used a 1.5km radius targeting office clusters and residential areas near Gulberg III. Weekend campaigns used a 3km radius for special-occasion diners willing to travel further.

Campaign architecture:

• Weekday lunch campaign: 1.5km radius around Gulberg III, dayparting 9am-2pm Tuesday-Thursday, age 25-45 (office workers, professionals), interest targeting for “food & dining,” “lunch,” “business networking,” and “local restaurants.” Ads featured weekday-specific offers (Weekday Platter PKR 1,499, Group of 4+ gets 15% off, Corporate catering available).
• Weekend dinner campaign: 3km radius, dayparting 5pm-10pm Friday-Saturday, age 22-50, broader interests including “date night,” “family dining,” and “special occasions.” Ads highlighted ambiance, weekend specials, and reservations.

We built creative variations testing Pakistan-culture-aware hooks against generic food photography. Offer creatives tested specific weekday deals (Weekday Executive Lunch, Corporate Group Packages, Tuesday Special). Urgency creatives tested limited-time language (Today Only, Valid 12pm-3pm, Weekday Exclusive). Social-proof creatives featured customer testimonials, Google Review ratings, and “Popular with nearby offices” messaging.

Landing page optimization focused on mobile-first experiences since 80%+ of Meta traffic was mobile. We built a simple mobile-optimized reservation page with: today’s availability, quick-view menu highlights, weekday offer banner, one-click reservation (phone or WhatsApp), and embedded location map. The page loaded in under 2 seconds on 4G and had clear calls-to-action for reservation (call now, WhatsApp us, book online).

For tracking, we implemented Meta Pixel events for page views, menu clicks, reservation clicks, and estimated conversions (correlated with reservation data). We set up custom conversions for “reservation initiated” and “reservation confirmed” to optimize campaigns toward actual bookings rather than just page visits. WhatsApp clicks were tracked as micro-conversions since many Pakistani customers prefer messaging for reservations.

Creative Test	Offer Message	Urgency Element	CTR	Reservation Conversion
Generic food photo	None	None	2.1%	3.2%
Weekday Platter offer	PKR 1,499 executive lunch	Valid Tuesday-Thursday	3.4%	6.8%
Group discount	15% off groups of 4+	Today only	4.2%	8.9%
Corporate catering	Office lunch packages	Call for availability	3.8%	7.4%
Social proof	”Popular with nearby offices”	Limited weekday seating	3.1%	5.6%

Phase 3 — Daypart optimization and offer scaling (Weeks 4-6)

With campaigns live, we moved into iterative optimization using weekly performance data. We tested multiple daypart schedules within the 9am-2pm window to identify peak conversion times. Data revealed that 10am-12pm produced the highest reservation rate (9.8%) as office workers planned lunch, while 12pm-2pm had higher engagement but lower conversion (many had already decided). We adjusted budget allocation to concentrate spend in the 10am-12pm window while maintaining lighter presence 12pm-2pm for same-day deciders.

Offer testing revealed that group discounts (15% off for 4+ people) outperformed individual offers by 2.3x on reservation conversion. Pakistani office-lunch culture often involves group dining, so creatives featuring team lunches, office celebrations, and colleague gatherings resonated. We scaled group-focused creatives and added corporate-catering messaging for larger groups (10+ people).

Audience testing showed that custom audiences of website visitors and page engagers produced lower cost per acquisition (CPA) but limited scale. Lookalike audiences (1% similarity) based on recent weekday customers produced the best balance of scale and efficiency. We excluded audiences who had already engaged with weekend campaigns to prevent cross-contamination.

We implemented frequency caps to prevent ad fatigue—limiting impressions to 3 per user per 7-day period. Data showed that performance declined after 4+ impressions as users became blind to the creatives. Frequency caps reduced wasted impressions while maintaining reach within the 1.5km radius.

Dayparting analysis revealed surprising patterns: Tuesday campaigns outperformed Thursday by 41% on ROAS, suggesting that mid-week had more office workers seeking lunch options compared to Thursday (closer to weekend when some employees work from home or leave early). We increased Tuesday budget by 35% and reduced Thursday spend accordingly, focusing on the strongest-performing weekday.

Remarketing campaigns launched in week 5, targeting website visitors who viewed the menu or made reservations but hadn’t visited recently. We created a “Come back this weekday” offer with a 10% discount code for returning customers. Remarketing contributed 8% of total weekday covers by week 6 with a 28% lower cost per cover than cold traffic.

Optimization Lever	Initial State	Optimized State	Impact
Daypart scheduling	9am-2pm flat	Concentrated 10am-12pm (70% of budget)	ROAS improved 22%
Offer type focus	Generic food photo	Group discount (15% off 4+ people)	Reservation conversion up 178%
Radius targeting	5km city-wide	1.5km hyperlocal for weekday lunch	Cost per cover down 41%
Audience strategy	Broad interests	Lookalike 1% based on weekday customers	CPA down 33%
Frequency management	No caps	3 impressions per 7 days	Ad fatigue reduced, engagement quality up

Phase 4 — Measurement and seasonal compounding (Weeks 6-8)

The final phase focused on measurement automation, seasonal planning, and scaling what worked. We built a custom dashboard connecting Meta Ads data, reservation system records, and POS transaction data to visualize the full funnel from impression to cover to revenue. The dashboard refreshed daily and showed metrics by daypart, creative, and geography—enabling data-driven decisions on budget allocation.

Revenue analysis revealed that weekday lunch customers had a 30-day repeat rate of 37% compared to 22% for weekend first-timers. Weekday customers were primarily local office workers and residents who returned regularly, making them higher lifetime value despite lower average check size. We adjusted bidding strategy to prioritize weekday acquisition even if upfront CPA appeared higher, knowing that repeat business would compound over months.

We tested seasonal variations—Ramadan awareness messaging (adjusting for fasting hours), monsoon-themed comfort food offers, and back-to-school September campaigns for nearby families. Ramadan campaigns paused during fasting hours and shifted to Iftar/dinner messaging. Monsoon campaigns featured indoor comfort food and “Rainy Day Lunch” offers. Seasonal relevance improved engagement by 15-20% compared to generic creatives.

Competitor monitoring revealed that other Gulberg restaurants began increasing their weekday ad spend in response to the restaurant’s success. We maintained differentiation by focusing on group offers and corporate-catering language that competitors weren’t emphasizing. We also tested Instagram-first campaigns with Reel-format video content showing behind-the-scenes kitchen preparation, which performed well with younger audiences (age 25-32).

By week 8, the restaurant achieved consistent weekday lunch occupancy of 65-70% (26 covers/day on average) compared to 35-40% at baseline. Cost per cover acquired dropped from PKR 145 to PKR 92 as campaigns optimized and remarketing audiences matured. ROAS on weekday campaigns improved from 1.7x to 3.4x as efficiency gains compounded. The restaurant increased weekday Meta budget to PKR 50,000 monthly while maintaining efficiency, knowing that incremental covers produced predictable revenue.

60-Day Outcome Metric	Baseline	Optimized	Change
Weekday lunch covers (Tue-Thu)	18/day	26/day	+44%
Weekday occupancy	38%	68%	+30 percentage points
Cost per cover acquired	PKR 145	PKR 92	-37%
Weekday campaign ROAS	1.7x	3.4x	+100%
Lunch-hour reservation conversion	4.8%	9.2%	+92%
Repeat-customer rate (30-day)	22%	37%	+68%
Average weekday check size	PKR 1,900	PKR 1,950	+3% (mix improvement)

Final Results

At 60 days, the Lahore restaurant’s Meta Ads program produced measurable, transformational outcomes for weekday lunch operations:

Metric	Before	After	Change
Weekday lunch covers per day (Tue-Thu)	18	26	+44%
Weekday lunch occupancy	38%	68%	+30 percentage points
Cost per cover acquired	PKR 145	PKR 92	-37%
Weekday campaign ROAS	1.7x	3.4x	+100%
Reservation conversion rate	4.8%	9.2%	+92%
Repeat-customer rate (30-day)	22%	37%	+68%
Monthly weekday Meta spend	PKR 45,000	PKR 50,000	+11% (intentional scale)
Estimated monthly revenue lift (weekday lunch)	—	PKR 156,000	New incremental revenue

What Made This Work

1. Location-first radius strategy—Rather than city-wide campaigns, we built hyperlocal 1.5km-radius campaigns targeting office clusters and residential areas near the restaurant. Data showed that 73% of weekday customers came from within 1.5km, yet previous campaigns spent 62% of budget on audiences >3km away. Radius targeting cut cost per cover by 41% and improved ROAS from 1.7x to 3.4x.

2. Daypart optimization for lunch-decision windows—Pakistani office workers decide lunch between 10am-1pm on the same day, often in group discussions around 11am-12pm. We shifted ad scheduling from evening hours (6pm-10pm) to 9am-2pm, concentrating 70% of budget in the 10am-12pm peak. This aligned campaigns with decision-making behavior and produced a 22% ROAS improvement.

3. Group-focused offer creatives—Testing revealed that group discounts (15% off for 4+ people) outperformed individual offers by 2.3x on conversion. Pakistani office-lunch culture heavily features group dining, so creatives featuring team lunches, office celebrations, and colleague gatherings resonated. We scaled group-focused messaging and added corporate-catering language.

4. Mobile-first reservation flows—80%+ of Meta traffic was mobile, so we built a mobile-optimized reservation page with quick-view menu, weekday offer banner, and one-click options (call, WhatsApp, book online). Reservation conversion improved from 4.8% to 9.2% as friction decreased and mobile experience improved.

5. Separate weekday and weekend tracks—We built distinct campaigns for weekday lunch and weekend dinner with different geographies, schedules, and creatives. This prevented cross-contamination (weekend messaging to weekday audiences) and allowed independent optimization. Weekday focused on office workers with group offers; weekend focused on special occasions with ambiance messaging.

6. Remarketing for customer retention—Remarketing campaigns targeting recent visitors with “Come back this weekday” offers produced incremental covers at 28% lower cost per cover than cold traffic. With a 37% 30-day repeat rate, weekday customers proved higher lifetime value than weekend first-timers, justifying investment in retention.

What Teams Can Apply

Pakistani restaurants can replicate this framework with these transferable takeaways:

1. Map your customer geography before targeting—Analyze your POS or reservation data to understand where your customers actually come from. Most weekday lunch customers are hyperlocal (within 1-2km), while weekend diners travel further. Build separate campaigns by geography and occasion rather than one generic city-wide campaign. Radius targeting dramatically improves efficiency.

2. Align ad scheduling with dining-decision windows—Pakistani lunch decisions happen same-day between 10am-1pm, often in group discussions. Schedule ads to run in this window (not evening hours when competition peaks) and concentrate budget in peak planning time (10am-12pm). Evening-heavy restaurant campaigns waste budget on audiences who have already decided.

3. Test group-focused offers for weekday lunch—Pakistani office-lunch culture heavily features group dining. Test offers like “15% off for groups of 4+,” “Team lunch package,” or “Corporate catering available” against generic food photography. Group-focused creatives typically outperform individual offers by 2-3x on conversion.

4. Build mobile-first reservation flows—Most Meta traffic is mobile, so ensure your reservation experience is optimized for small screens. Quick-view menu highlights, clear weekday offers, and one-click options (call, WhatsApp) reduce friction and improve conversion. Avoid complex multi-step forms on mobile.

5. Separate weekday and weekend strategies—Weekday and weekend restaurant audiences differ significantly in behavior, geography, and motivation. Build distinct campaigns with tailored messaging, scheduling, and optimization. Weekday focuses on convenience and value for office workers; weekend focuses on ambiance and experience for special occasions.