Half Your Marketing Budget Is Working - You Just Don't Know Which Half

Six Promotions, One Question Nobody Could Answer

Hana is the marketing director for a 16-location restaurant group across Riyadh and Jeddah. On any given week, she's running six concurrent promotions: a 30% discount on Entertainer, a "Buy 1 Get 1" on a new menu item through the brand's app, an influencer partnership driving traffic to three locations, a Ramadan set menu campaign, a corporate catering promotion, and a loyalty program double-points week.

Total monthly marketing spend: SAR 320,000.

At the end of each month, Hana reports to the CEO. Revenue is up 8% month over month. Marketing spend is SAR 320,000. The CEO asks a simple question: "Which of these six campaigns is actually working?"

Hana doesn't know. Not because she's not trying - her team tracks redemption counts, coupon usage, social engagement, and every metric the platforms provide. The problem is that none of these metrics answer the CEO's actual question, which is really: "Which campaigns are generating incremental revenue that wouldn't have happened without the promotion - and which are just subsidizing guests who would have come anyway?"

Redemption counts look great on the Entertainer promotion: 2,400 redemptions last month. But how many of those 2,400 guests would have dined at the restaurant at full price without the Entertainer offer? If the answer is 1,800 (which Sundae's analysis eventually revealed), then the promotion didn't drive 2,400 incremental visits. It drove 600 incremental visits and gave a 30% discount to 1,800 guests who didn't need one.

The Buy 1 Get 1 promotion had lower redemptions - only 580. But because it was tied to a new menu item that guests hadn't tried before, 72% of redemptions came from genuinely incremental visits. At SAR 95 average check, 580 redemptions at 72% incrementality generated SAR 39,700 in incremental revenue - 4.1x the return of the Entertainer promotion per SAR spent, despite having one-quarter the redemption count.

Hana had been allocating 70% of her promotional budget to discount-based campaigns because the redemption numbers looked bigger. The actual incremental ROI said she should have been doing the opposite.

This is the problem Sundae's Marketing Performance Intelligence solves.

The Attribution Gap in Restaurant Marketing

Restaurant marketing suffers from a fundamental attribution problem that most other industries solved years ago. E-commerce tracks every click from ad impression to purchase. SaaS platforms attribute signups to specific campaigns with precision. Restaurants? Restaurants know that someone used a coupon - but not whether that person would have walked in anyway.

1. Correlation vs. Causation

Revenue went up during the campaign period. But revenue also goes up during cooler weather, holiday seasons, new menu launches, and after positive social media moments. Without controlled measurement, it's impossible to separate campaign impact from background demand fluctuations.

2. Redemption vs. Incrementality

Redemption count measures how many people used a promotion. Incrementality measures how many of those people wouldn't have visited without the promotion. They answer different questions, but most restaurant marketing teams report redemptions because incrementality is harder to measure. That leaves campaigns that mainly subsidize existing demand looking stronger than they really are.

3. Cannibalization Blindness

When a 30% discount promotion runs, it doesn't just attract new guests - it also attracts existing guests who trade down from full price to discounted price. This cannibalization is invisible in campaign reporting because the discounted transaction gets recorded as a "campaign success" even when it simply reduced margin on a visit that would have happened anyway.

4. Cross-Campaign Interference

With multiple promotions running simultaneously, it's impossible to attribute a transaction to a single campaign. A guest who saw an Instagram ad, received a loyalty email, and noticed an Entertainer offer - which campaign drove the visit? Traditional reporting counts it as a "win" for whichever promotion the guest happened to redeem, regardless of which one actually influenced the decision.

5. Time-Horizon Mismatch

Marketing campaigns have different time horizons. A discount promotion generates immediate transactions but may not build lasting behavior. An influencer partnership may generate social buzz that converts to visits over weeks or months. Brand-building content may not generate measurable transaction impact for quarters. Comparing these on a monthly ROI basis penalizes long-term investments and rewards short-term subsidies.

What Sundae's Marketing Performance Intelligence Does

Campaign-to-Transaction Attribution

Sundae connects marketing campaign data to POS transaction data at the guest level, enabling true attribution:

• Direct attribution: Guest redeemed a specific offer → transaction is attributed to that campaign
• Influenced attribution: Guest was exposed to campaign (received email, saw social post, was in geo-targeted zone) and visited within the attribution window → transaction is partially attributed to the campaign
• Baseline comparison: Guest's visit pattern is compared to their pre-campaign pattern to determine whether the visit was incremental or would have occurred regardless

This three-layer attribution model produces a materially different picture than simple redemption counting. Across Sundae's customer base, the average campaign shows 40-60% incrementality - meaning 40-60% of redemptions represent genuinely new visits, and the rest are subsidies to existing demand.

Incrementality Scoring

For every campaign, Sundae calculates an incrementality score: the percentage of attributed transactions that represent genuinely incremental revenue. This score is the single most important metric in campaign evaluation:

• High incrementality (70%+): Campaign is effectively driving new demand. Scale it.
• Medium incrementality (40-70%): Campaign drives some new demand but also subsidizes existing. Optimize targeting to reduce subsidy.
• Low incrementality (<40%): Campaign is primarily subsidizing existing demand. Restructure or discontinue.

The incrementality score transforms marketing evaluation from "did we get redemptions?" to "did we get revenue we wouldn't have gotten otherwise?" - which is the only question that matters for ROI.

Promotional Cannibalization Detection

Sundae identifies when promotions cannibalize full-price revenue:

• Price cannibalization: Guests who regularly visit at full price switch to discounted visits during promotion periods. Net impact: negative margin on the same guest.
• Timing cannibalization: Guests shift their visit timing to align with promotional periods rather than increasing total visits. Monthly visit count stays the same; promotional cost increases.
• Menu cannibalization: Promotions on specific items draw orders away from higher-margin alternatives. Total covers may increase, but margin per cover decreases.

Cannibalization analysis reveals the true cost of promotions. A campaign that generates SAR 100,000 in attributed revenue but cannibalizes SAR 65,000 in full-price revenue has a net incremental impact of SAR 35,000 - a very different number than what the campaign report shows.

Discount Depth Analysis

Not all discounts are created equal. Sundae analyzes the relationship between discount depth and incremental response:

• 10% discount: Minimal incremental impact - most guests who redeem would have visited anyway
• 20% discount: Moderate incremental impact - begins to attract price-sensitive guests who wouldn't otherwise visit
• 30% discount: Higher incremental guest count but significant cannibalization of full-price guests
• 50%+ discount (BOGO): Highest incremental guest count but lowest per-guest margin; effective for trial, destructive for margin if sustained

This analysis helps operators find the optimal discount depth: deep enough to drive genuine incrementality, shallow enough to minimize cannibalization. For most casual dining concepts in GCC markets, the optimal discount depth for incremental guest acquisition is 20-25% - enough to motivate trial without training guests to wait for discounts.

Channel Mix Optimization

Sundae evaluates marketing channel effectiveness by connecting spend to incremental revenue:

• Social media advertising: Cost per incremental visit, by platform (Instagram, TikTok, Snapchat, X)
• Influencer partnerships: Incremental visits attributed to influencer content, measured against the timing and geography of the content
• Platform promotions (Entertainer, Zomato Gold, etc.): True incremental revenue vs. subsidized revenue
• Loyalty program campaigns: Re-engagement effectiveness of loyalty communications by segment
• Corporate/B2B outreach: Catering and event revenue generated by sales efforts
• Direct marketing (email, SMS, push): Open-to-visit conversion rates by message type and segment

Each channel receives an incremental cost-per-acquisition (iCPA) score - the cost to generate one genuinely incremental visit. This enables apples-to-apples comparison across channels that produce very different volumes and very different metrics.

Campaign Experimentation Framework

For operators who want to move beyond measurement to optimization, Sundae supports structured experimentation:

• A/B offer testing: Run two different offers to similar audience segments and measure incremental response
• Control group holdouts: Exclude a random subset of eligible guests from a campaign to measure true lift against a control
• Geographic testing: Run a campaign at half your locations and use the other half as a control group
• Sequential testing: Test one campaign element at a time (discount depth, creative, channel, timing) to isolate what drives response

This framework transforms marketing from a creative exercise into a data-driven optimization process. Each campaign generates learning that improves the next campaign's targeting, offer design, and channel allocation.

Building a Marketing Intelligence Practice

Step 1: Establish Baseline Demand

Before evaluating any campaign, you need to know what "normal" looks like. Sundae establishes demand baselines by location, day of week, daypart, and season. Any campaign evaluation is measured against this baseline - not against last month or last year, but against the expected demand for that specific period.

Step 2: Define Incrementality Thresholds

Not every campaign needs to hit 70% incrementality. Brand-building campaigns and loyalty retention campaigns serve different purposes than guest acquisition campaigns. Define acceptable incrementality thresholds by campaign objective:

• Acquisition campaigns: 50%+ incrementality target
• Trial/new menu campaigns: 60%+ incrementality target (these should primarily reach new-to-item guests)
• Retention campaigns: Measured by retention lift rather than incrementality (did at-risk guests come back?)
• Brand campaigns: Measured by awareness and consideration metrics over longer time horizons

Step 3: Rationalize Campaign Portfolio

Most restaurant marketing teams run too many simultaneous campaigns with too little measurement. Sundae's Marketing Intelligence helps rationalize the portfolio:

• Kill low-incrementality campaigns that are primarily subsidizing existing demand
• Scale high-incrementality campaigns that are demonstrably driving new revenue
• Test uncertain campaigns using control groups before committing full budget
• Rebalance channel mix based on incremental cost-per-acquisition, not redemption volume

A typical marketing portfolio optimization identifies 25-40% of spend that can be reallocated from low-performing to high-performing campaigns - improving total incremental revenue by 30-50% without increasing total marketing budget.

Step 4: Implement Continuous Learning

Marketing intelligence is not a one-time audit. It's an ongoing practice:

• Weekly: Review active campaign incrementality scores. Flag any campaign with declining incrementality (common with discount campaigns that train guests to wait for offers).
• Monthly: Full campaign portfolio review. Reallocate budget from underperformers to outperformers.
• Quarterly: Strategic channel mix review. Evaluate emerging channels and test new formats.
• Annually: Full marketing strategy review informed by 12 months of incrementality data.

The GCC Marketing Context

GCC restaurant markets have characteristics that make marketing intelligence particularly critical:

Promotion-heavy culture: GCC diners are sophisticated promotion users. Platforms like Entertainer and Zomato Gold have conditioned a significant segment of the dining market to seek offers. This means discount promotions have higher redemption rates but also higher cannibalization - more guests who would have visited anyway are using the discount as a bonus rather than an incentive.

Influencer-driven discovery: Influencer marketing drives a larger share of restaurant discovery in GCC markets than in most global markets. But influencer economics are opaque - most operators pay per post without understanding cost per incremental visit. Sundae's attribution helps operators identify which influencer partnerships actually drive visits, not just engagement.

Multi-platform presence: GCC restaurant groups typically maintain presence on multiple deal and booking platforms simultaneously. Without cross-platform incrementality analysis, operators can't distinguish between platforms that drive new demand and platforms that merely redirect existing demand.

Seasonal demand swings: Ramadan, summer heat, holiday seasons, and major events create dramatic demand fluctuations. Campaigns running during demand peaks may appear successful simply because of background demand, while campaigns running during troughs may appear unsuccessful despite driving genuine incrementality.

What Operators Should Do This Week

Action 1: Pull redemption data for your three highest-spend campaigns. For each one, estimate what percentage of redeemers would have visited at full price without the promotion. Be honest - the number is almost always higher than you think.

Action 2: Calculate your true incremental cost per acquisition. Take total campaign spend, divide by the number of genuinely incremental visits (not total redemptions). Compare this across campaigns. You'll likely find a 3-5x variation in efficiency.

Action 3: Identify your highest cannibalization risk. Which promotion is most likely being used by guests who would have come anyway? That's your first optimization target.

Action 4: Stop reporting redemption counts to leadership. Start reporting incremental revenue and incrementality scores. This changes the conversation from "did we get lots of people to use the offer?" to "did we generate revenue that wouldn't have happened otherwise?"

Closing and Call to Action

The famous quote about half of advertising being wasted came from an era when measurement was hard. In restaurant marketing today, the data usually exists - it just has not been connected. POS transactions, campaign exposure records, guest visit histories, and platform redemption data can be tied together to show which spend is creating new demand and which spend is discounting demand you already had.

Sundae's Marketing Performance Intelligence makes this connection. It transforms marketing from a cost center that reports redemption counts into a revenue driver that reports incremental ROI - giving marketing directors the data to defend what works, cut what doesn't, and reallocate budget where it generates real growth.

Book a demo to see Sundae's Marketing Performance Intelligence on your own campaign data - and find out which half of your marketing budget is actually working.