A recent campaign aimed at driving EWR ordering for a regional food delivery service illustrated the critical interplay between precise targeting and creative messaging in paid ad conversion funnels. This teardown unpacks how a substantial budget translated into tangible results, and where the strategy encountered unexpected friction.
Key Takeaways
- A $150,000 budget over six weeks yielded a 3.5x ROAS for EWR ordering, demonstrating the potential of focused paid ad campaigns.
- Precise geo-fencing around Newark Liberty International Airport (EWR) combined with interest-based targeting significantly reduced Cost Per Lead (CPL) to $8.50.
- Dynamic creative optimization, specifically A/B testing video versus static image ads, revealed video ads generated a 25% higher Click-Through Rate (CTR).
- Initial conversion rates were hindered by a convoluted checkout process, which required a 40% reduction in form fields to improve.
- Continuous monitoring of real-time performance metrics on platforms like Google Ads and Meta Business Suite allowed for daily budget reallocation and creative adjustments.
“Growth marketing teams need to connect AEO to acquisition metrics quickly enough to act on the signal and justify investment. The teams doing that now are building a playbook that will be much harder to replicate once the channel matures.”
Campaign Overview: Driving EWR Ordering for “FlyBites”
Our objective was straightforward: increase direct EWR ordering for FlyBites, a new pre-order meal service targeting travelers and airport staff at Newark Liberty International Airport. The campaign ran for six weeks, from September 1 to October 15, 2026, with a total budget of $150,000. We focused primarily on Google Search Ads and Meta (Facebook/Instagram) platforms, using their granular targeting capabilities.
Initial Strategy: Pinpointing the EWR Audience
The core strategy hinged on reaching individuals with a high likelihood of needing airport food services. This meant a dual approach: geographic precision and behavioral inference. For Google Search, we targeted keywords like “EWR food delivery,” “Newark airport meals,” “pre-order EWR,” and specific airline lounge names within a 5-mile radius of the airport. This radius captured both travelers arriving/departing and the significant workforce operating within the airport complex. On Meta, our targeting expanded to include interests such as “frequent flyer programs,” “business travel,” “airport lounges,” and “specific airlines operating out of EWR.” We also uploaded a custom audience of known airport employees from a previous partnership, ensuring we weren’t just guessing at their presence.
Our initial hypothesis was that convenience and speed would be the primary motivators. The creative reflected this, emphasizing “Skip the line” and “Fresh meals delivered to your gate.”
Creative Approach: Video Versus Static
We launched with a mix of creative assets across both platforms. For Google Display Network and Meta, we developed a series of short, 15-second video ads showing the ease of ordering and the quality of the food. These were complemented by static image ads featuring appetizing meal shots and clear calls to action. A/B testing was baked into the campaign from day one, allowing us to compare performance metrics between video and static formats. The videos consistently outperformed static images, generating a 25% higher Click-Through Rate (CTR) on average, settling at 1.8% for video versus 1.4% for static on Meta. This initial data quickly informed a reallocation of 60% of our display budget towards video assets.
Targeting Refinements: From Broad to Hyper-Specific
While the initial geo-fencing was effective, we noticed a higher Cost Per Conversion (CPC) for some broader keywords. We refined our Google Search targeting to focus on long-tail keywords that indicated stronger purchase intent, such as “EWR terminal B food pre-order” or “healthy airport meals Newark.” This move, implemented in week three, saw our average Cost Per Click (CPC) drop from $2.10 to $1.65 for these refined segments. For Meta, we began layering exclusion audiences, removing individuals who had recently interacted with competing food delivery services not specific to airports, further tightening our focus. This iterative process is important. The market doesn’t stand still, and neither should your targeting strategy.
Conversion Funnel Analysis: Identifying Friction Points
The conversion funnel for EWR ordering was designed as follows:
- Ad Impression
- Click to Landing Page
- Meal Selection
- Customization/Add-ons
- Checkout Form Completion
- Payment
- Order Confirmation
Our initial landing page, while visually appealing, had a relatively high bounce rate of 45%. User session recordings and heatmaps from tools like Hotjar revealed that users were dropping off primarily at the checkout form. The form required 12 distinct data fields, including flight number, gate number, and a detailed delivery time window. This was simply too much friction for a mobile-first audience in a hurry. A eMarketer report from late 2025 highlighted that reducing form fields can increase mobile conversion rates by up to 20%. We took that to heart.
Optimization Steps: Simplifying the Path to Purchase
Based on the funnel analysis, we implemented several key optimizations:
- Checkout Form Reduction: We reduced the checkout form from 12 fields to 7 essential fields, making gate and flight number optional for initial order placement and prompting for it post-purchase via SMS. This change alone saw a 30% increase in form completion rates.
- Mobile Responsiveness: While the site was technically responsive, we optimized load times for mobile devices, particularly for users on airport Wi-Fi, which can be inconsistent. Page load speeds decreased by an average of 1.5 seconds.
- Dynamic Pricing Promos: We introduced limited-time offers, such as “10% off your first EWR order” for new customers, dynamically served to users who had visited the site but not completed a purchase. These retargeting ads achieved a 3.2% CTR and contributed significantly to closing hesitant users.
What Worked and What Didn’t: Key Learnings
What Worked:
- Hyper-Localized Targeting: The combination of geo-fencing and interest-based targeting around EWR proved highly effective. Our Cost Per Lead (CPL) for a qualified customer (defined as someone who added an item to their cart) was a respectable $8.50.
- Video Creative: The dynamic and engaging nature of video ads resonated more with our target audience, leading to higher engagement rates.
- Iterative Optimization: The willingness to make rapid changes based on real-time data, particularly regarding the checkout process, was paramount. We didn’t wait for the campaign to finish to make significant adjustments.
- Retargeting: Specific retargeting campaigns for cart abandoners and landing page visitors who didn’t convert yielded a strong Return on Ad Spend (ROAS) of 4.1x for those segments.
What Didn’t Work as Expected:
- Broad Keyword Matching: Initially, some broader keywords like “airport food” attracted clicks but few conversions, leading to wasted spend. Refinement was critical.
- Complex Checkout: The initial checkout process was a significant barrier. We overestimated users’ willingness to input detailed information upfront. This was a hard lesson in user experience over perceived data collection needs.
- Lack of Real-time Inventory Integration: While not strictly an ad funnel issue, the absence of real-time inventory updates on the ordering platform sometimes led to customers ordering unavailable items, resulting in cancellations and a poor customer experience. This indirectly impacted ad performance by reducing repeat purchases.
Performance Metrics and ROAS
Over the six-week period, the campaign achieved the following:
- Total Impressions: 1.8 million
- Total Clicks: 42,000
- Overall CTR: 2.3%
- Total Conversions (Completed Orders): 6,500
- Average Cost Per Conversion: $23.08
- Average Order Value (AOV): $80.00
- Total Revenue Generated: $520,000
- Return on Ad Spend (ROAS): 3.5x
This 3.5x ROAS indicates that for every dollar spent on advertising, FlyBites generated $3.50 in revenue. While initial projections aimed for 4x, the early checkout friction pulled this down slightly. However, the optimizations helped us recover and stabilize performance in the latter half of the campaign. The conversion rate from click to order was 15.5%, which, while not industry-leading, was a substantial improvement from the initial 10% we saw in the first two weeks.
Regular performance reviews, often daily during the initial launch phase, allowed us to pivot quickly. We leveraged automated rules within Google Ads to adjust bids for high-performing keywords and paused underperforming ad groups entirely. This proactive management saved a considerable portion of the budget from being misspent.
Conclusion
The FlyBites EWR ordering campaign underscored that even with a strong budget, continuous optimization and a deep understanding of the customer journey are non-negotiable. Focus on removing friction points in your conversion funnel, especially on mobile, and be prepared to adapt your strategy based on real-time data. That flexibility is often the difference between moderate success and significant growth.
What is a good ROAS for paid ad campaigns?
A generally accepted good Return on Ad Spend (ROAS) for most businesses is 3:1 or 4:1, meaning you generate $3 to $4 in revenue for every $1 spent on advertising. However, this can vary significantly by industry, profit margins, and business goals, with some businesses targeting a lower ROAS for brand awareness or a much higher one for direct-response campaigns.
How often should I review my paid ad campaign performance?
During the initial launch phase of a new campaign, daily reviews are advisable to identify immediate issues and opportunities. Once a campaign stabilizes, weekly or bi-weekly detailed reviews are typically sufficient, with daily checks on key metrics like spend and conversions to catch any sudden anomalies.
What are common reasons for high bounce rates on landing pages?
High bounce rates often stem from a mismatch between the ad’s promise and the landing page’s content, slow page load times, poor mobile responsiveness, confusing navigation, or an overwhelming amount of information or form fields. Ensuring a clear, concise, and fast user experience is key.
Can geo-fencing be too restrictive for ad campaigns?
While precise geo-fencing can be highly effective for local businesses or specific locations like airports, it can also be too restrictive if your target audience extends beyond that immediate area. It’s important to balance precision with potential reach, often by starting narrow and gradually expanding if performance allows.
Is it better to use video or static images in paid ads?
Neither video nor static images are universally “better”. Performance depends on the platform, audience, and campaign objective. However, video often achieves higher engagement and CTR due to its dynamic nature. A/B testing both formats is the most reliable way to determine what resonates best with your specific audience.
