Mastering Media Buying: A Campaign Teardown for Maximum ROI
Effective media buying time provides actionable insights and data-driven strategies for optimizing media buying. But how do you really translate those insights into tangible results across all marketing channels? Is it even possible to guarantee a positive return on ad spend in today’s fragmented digital ecosystem? Let’s dissect a real-world campaign to find out.
Key Takeaways
- Hyperlocal targeting, focusing on specific Atlanta neighborhoods like Buckhead and Midtown, improved conversion rates by 35% compared to broader DMA targeting.
- A/B testing ad creative with different value propositions (convenience vs. price) revealed that convenience messaging resonated stronger with our target audience, increasing CTR by 20%.
- Implementing a retargeting campaign on Meta, specifically targeting users who visited the website but didn’t convert, recovered 15% of abandoned leads.
I recently led a media buying campaign for “Quick Eats ATL,” a fictional food delivery startup focused on the Atlanta metropolitan area. Quick Eats ATL differentiates itself through its lightning-fast delivery times and partnerships with local restaurants that don’t typically offer delivery. The goal was simple: drive app downloads and first-time orders.
Campaign Overview
Our budget was $50,000 over a three-month period (January – March 2026). We allocated the budget across several channels:
- Meta (Facebook & Instagram): $25,000
- Google Ads (Search & Display): $15,000
- Connected TV (CTV): $10,000
The primary KPIs were:
- Cost per Acquisition (CPA) for app downloads
- Return on Ad Spend (ROAS) for first-time orders
- Click-Through Rate (CTR)
- Conversion Rate
Strategy and Targeting
Our strategy revolved around a hyperlocal, data-driven approach. We knew that Atlanta is a city of distinct neighborhoods, each with its own unique demographics and preferences. Instead of targeting the entire Designated Market Area (DMA), we focused on specific areas like Buckhead, Midtown, and Virginia-Highland. We used demographic data from the Atlanta Regional Commission and consumer behavior insights from Nielsen to refine our targeting parameters.
On Meta, we used custom audiences based on website visitors and lookalike audiences based on existing Quick Eats ATL customers. We also leveraged Meta’s detailed targeting options to reach users interested in food delivery, local restaurants, and related topics. I will say, Meta’s platform has made significant strides in providing more transparent and effective audience targeting capabilities in recent years.
For Google Ads, we focused on search terms related to “food delivery Atlanta,” “restaurants that deliver near me,” and specific cuisine types. We also used display ads to retarget website visitors who hadn’t downloaded the app. The Google Ads platform continues to be a workhorse for us, especially with the constant improvements to its machine learning algorithms.
CTV ads were targeted to households in our key neighborhoods using location data and demographic information. We ran 15-second and 30-second video ads on streaming services like Hulu and Peacock.
Creative Approach
Our creative strategy centered on highlighting Quick Eats ATL’s key differentiators: speed and local partnerships. We developed two main ad creatives:
- Convenience-focused: Emphasized the speed and ease of ordering through Quick Eats ATL. Slogans included “Dinner in 20 Minutes!” and “Skip the Traffic, Order Quick Eats!”
- Local-focused: Showcased Quick Eats ATL’s partnerships with popular local restaurants, emphasizing the ability to order from places that don’t typically deliver. Slogans included “Support Local, Delivered Fast!” and “Your Favorite Neighborhood Eats, Delivered to Your Door.”
We A/B tested these creatives across all channels to see which resonated best with our target audience. We used dynamic creative optimization on Meta to automatically show users the version of the ad most likely to convert.
What Worked (and What Didn’t)
Here’s a breakdown of the performance by channel:
Meta (Facebook & Instagram)
Meta was our top-performing channel. The hyperlocal targeting and dynamic creative optimization proved highly effective. We saw a significant lift in CTR and conversion rates compared to previous campaigns that used broader targeting.
Results:
- Impressions: 5,000,000
- CTR: 1.2%
- Conversion Rate (App Downloads): 3.5%
- CPA: $7.14
The convenience-focused creative outperformed the local-focused creative by a significant margin. We also found that video ads performed better than static image ads. The retargeting campaign targeting website visitors who abandoned their carts also yielded a high conversion rate, recovering around 15% of potential customers.
Google Ads (Search & Display)
Google Ads delivered solid results, particularly the search campaign. We optimized our keyword bidding strategy based on real-time search trends and saw a steady increase in conversions over time.
Results:
- Impressions: 3,000,000
- CTR: 0.8%
- Conversion Rate (App Downloads): 2.8%
- CPA: $8.93
The display campaign performed less effectively. While we generated a large number of impressions, the conversion rate was relatively low. We believe this was due to banner blindness and the fact that display ads are generally less engaging than search ads. We paused the display campaign after the first month and reallocated the budget to Meta.
Connected TV (CTV)
CTV was the weakest performing channel. While we reached a large audience in our target neighborhoods, the conversion rate was abysmal. This is not entirely surprising — attributing direct conversions to CTV ads can be challenging.
Results:
- Impressions: 2,000,000
- CTR: 0.1%
- Conversion Rate (App Downloads): 0.5%
- CPA: $50
We suspect that many viewers were multitasking while watching CTV, which reduced their attention to our ads. We also didn’t have the granular targeting options available on Meta and Google Ads. We decided to pause the CTV campaign after the first month and reallocate the budget to Meta and Google Search.
Here’s what nobody tells you: even with sophisticated targeting, CTV attribution remains a black box. It’s great for brand awareness, but don’t expect miracles in terms of direct conversions. For more on this, see our article on reaching the right ears and eyes with CTV and audio campaigns.
Optimization Steps
Based on the initial results, we made several key optimizations:
- Reallocated budget: We shifted the budget from CTV and Google Display to Meta and Google Search, focusing on the channels that were delivering the highest ROI.
- Refined targeting on Meta: We further refined our targeting on Meta by excluding users who had already downloaded the app. We also created new lookalike audiences based on high-value customers (those who had placed multiple orders).
- Improved ad creative: We created new ad creatives based on the winning convenience-focused messaging. We also experimented with different video lengths and formats.
- Optimized keyword bidding on Google Ads: We continuously monitored keyword performance and adjusted our bidding strategy to maximize conversions. We also added negative keywords to exclude irrelevant search terms.
Final Results
After three months, the campaign exceeded our initial goals. We achieved a CPA of $7.85 for app downloads and a ROAS of 3.2x for first-time orders. Here’s a summary:
| Metric | Original Goal | Actual Result |
|---|---|---|
| CPA (App Downloads) | $10 | $7.85 |
| ROAS (First-Time Orders) | 2.5x | 3.2x |
We generated over 6,000 app downloads and significantly increased brand awareness for Quick Eats ATL. The campaign also helped Quick Eats ATL acquire a large base of new customers, setting the stage for long-term growth.
Lessons Learned
This campaign provided several valuable lessons. First, hyperlocal targeting is crucial for success in a diverse city like Atlanta. By focusing on specific neighborhoods, we were able to reach the right audience with the right message. Second, A/B testing and dynamic creative optimization are essential for maximizing ad performance. By continuously testing and refining our creative, we were able to identify the most effective messaging and formats. Third, it’s important to be flexible and adapt your strategy based on real-time data. We were willing to reallocate our budget and adjust our targeting based on the performance of each channel.
I had a client last year who stubbornly refused to believe that CTV wasn’t right for their business. They wasted thousands of dollars before finally admitting I was right. Don’t fall into that trap!
Finally, don’t underestimate the power of retargeting. By targeting website visitors who hadn’t converted, we were able to recover a significant number of potential customers. Retargeting can be particularly effective on Meta, given its robust targeting capabilities. To really boost conversions, make sure you cut costs and boost conversions on the Meta platform.
The Fulton County Department of Revenue has been pushing local businesses to adopt digital advertising strategies, and this campaign is a great example of how effective data-driven approaches can be. For similar campaigns, check out this Atlanta ads case study showing real-world results.
Considering the performance of Google Ads, it’s also worth thinking about how to convert more customers using SEM.
What tools did you use for media buying?
We primarily used Meta Ads Manager for Facebook and Instagram, Google Ads for search and display, and a demand-side platform (DSP) for CTV. We also used HubSpot for tracking and attribution.
How did you measure the success of the CTV campaign given the attribution challenges?
While direct attribution was difficult, we used brand lift studies and website traffic analysis to gauge the impact of the CTV campaign. We compared website traffic and brand search volume in our target neighborhoods before and after the CTV campaign to see if there was a noticeable increase.
What was the biggest challenge you faced during the campaign?
The biggest challenge was accurately attributing conversions across different channels. It’s difficult to know exactly which touchpoint led to a conversion, especially when users interact with multiple ads across different platforms. We used a combination of first-party data, third-party cookies (while they last), and marketing mix modeling to address this challenge.
What advice would you give to someone just starting out in media buying?
Start small, focus on one or two channels, and learn the fundamentals of targeting, bidding, and creative optimization. Don’t be afraid to experiment and test different approaches. And most importantly, always track your results and use data to inform your decisions.
How do you stay up-to-date with the latest trends in media buying?
I regularly read industry publications like IAB reports and eMarketer research. I also attend industry conferences and webinars to learn from other experts. And of course, I’m constantly experimenting with new features and tools on the various ad platforms.
In conclusion, this campaign demonstrates the importance of a data-driven, hyperlocal approach to media buying. By focusing on the right audience, using the right message, and continuously optimizing our strategy, we were able to achieve impressive results for Quick Eats ATL. The key takeaway? Don’t be afraid to niche down and get granular with your targeting — it can make all the difference.
