Optimizing TV and radio media buying using SmartSpot360 on AWS

This blog post was co-authored by Jhon Valencia (SVP of Technology & Data), Preston Porter (Vice President of Operations), and Claire McCue (Director of Marketing) at BMG360.

When you depend on high-volume lead generation, media buying efficiency directly impacts revenue and scalability. However, as campaigns expand across channels, optimizing traditional media (like TV and radio) becomes increasingly complex. You need a reliable ability to adjust placements in real time and predict campaign performance using live data.

To address this gap, BMG360 developed SmartSpot360™, an AI solution built on Amazon Web Services (AWS). SmartSpot360 transforms linear media buying by applying real-time analytics and predictive modeling to deliver optimized placements at scale.

We will share how you can modernize legacy media operations, improve efficiency, and enhance campaign performance using an AWS-powered machine learning (ML) platform.

Challenges and business context

Linear TV and radio advertising remain valuable channels for reaching broad audiences, but they’ve historically relied on manual processes and fixed decision rules. These methods worked but didn’t scale easily—particularly when trying to optimize spend in real time or predict performance across shifting audience behaviors and market conditions. You couldn’t readily adjust campaigns in real time or effectively apply predictive insights to future placements at speed and at scale.

BMG360 identified three core challenges in the traditional model:

1. Limited scalability of manual buying decisions
2. Lack of real-time adaptability to changing performance data
3. No predictive intelligence to guide future media investments

Before developing a proprietary solution, BMG360 evaluated multiple approaches:

• Manual optimization, based on historical data and planner expertise, often required significant time and introduced inefficiencies.
• Third-party media buying platforms, which offered a degree of automation, lacked the flexibility to integrate with proprietary datasets or adapt to unique campaign logic deployed by BMG360.
• Traditional business intelligence tools supported retrospective analysis but provided no mechanisms for real-time optimization or predictive modeling.

To address these constraints, BMG360 saw a clear opportunity: build an AI-driven system capable of learning from past performance, optimizing active campaigns in real time, and dynamically allocating spend across variables (such as daypart, rate, and creative mix).

The result was SmartSpot360: a machine learning-powered platform designed to bring scale, speed, and precision to traditional media buying.

How BMG360 engaged AWS

The SmartSpot360 architecture is designed to be use case–dependent, meaning its configuration and service stack may vary depending on campaign scale, data sources, and optimization objectives. The following overview represents one reference implementation designed for high-volume TV and radio use cases.

To support the development and scaling of SmartSpot360, BMG360 worked closely with AWS to evaluate architecture options for data processing, machine learning (ML) model training, and real-time API delivery. The AWS team provided guidance on selecting the right mix of serverless and ML services to balance performance, cost, and scalability.

BMG360 was able to quickly scale and find efficiency by engaging with a dedicated AWS team, as well as through a referred AWS Partner, Usage.AI, to verify a stable and long-term solution was implemented from the ground up. As the application’s potential was identified, several meetings to address resilience, efficiency, and Amazon SageMaker best practices culminated in a successful launch and expanded use case for the product.

The business continues to engage with the AWS team, including through on-site discovery sessions with AWS thought leadership. These sessions confirm both the AWS-powered solutions and overall cloud application strategies of BMG360 remain in line with or ahead of industry best practices.

Solution overview

To bring scale and intelligence to traditional media, SmartSpot360 optimizes TV and radio placements in real time. Built to process and analyze over 3,500 ad spots each week, SmartSpot360 continuously learns from performance data to refine buying decisions across a range of variables.

The system evaluates:

• Spot cost, length, and frequency
• Time-of-day and day-of-week trends
• Creative variation and effectiveness
• Seasonal and holiday-driven audience behavior

Using advanced predictive modeling, SmartSpot360 dynamically adjusts placements based on historical outcomes and live campaign performance. Instead of relying on fixed schedules or manual rule sets, the system allocates spend where it’s most likely to generate value. It automates what previously required extensive manual input and post-campaign analysis.

This approach improves operational efficiency, while increasing the likelihood of higher-quality leads and stronger conversion outcomes. By shifting optimization from reactive to proactive, SmartSpot360 enables data-informed decision-making at scale across thousands of placements and changing market conditions.

Figure 1: Automated retraining workflow.

Figure 2: Recommendation workflow.

The SmartSpot360 architecture is powered by AWS services that support high availability, scalability, and real-time performance:

• Amazon SageMaker is used to train, validate, and deploy machine learning models that predict spot performance and optimize allocation logic.
• AWS Glue handles extract, transform, and load (ETL) operations, ingesting performance data from multiple sources and preparing it for modeling.
• AWS Lambda executes real-time logic for budget recommendations, creative rotation, and time-slot selection without managing infrastructure.
• Amazon API Gateway facilitates secure, API-driven interactions with SmartSpot360, enabling seamless integration with planning and reporting tools.
• Snowflake on AWS serves as the centralized data warehouse, supporting large-scale queries across historical and in-flight campaign data.
• Amazon Simple Notification Service (Amazon SNS) coordinates event-driven communication between services by publishing messages to topics that trigger downstream processes (such as model retraining, data processing, or API updates).

With this architecture, BMG360 can continuously improve the performance, reduce operational complexity, and rapidly scale SmartSpot360 as new use cases emerge.

Business outcomes and metrics

Adopting AI-driven media optimization produces measurable improvements across efficiency, spend, and revenue.

Results achieved by BMG360, using SmartSpot360, during a health insurance TV campaign:

• Media spend decreased by 32%
• Revenue generated increased by 52%
• Return on investment (ROI) for every TV dollar doubled

Operationally, automating spot selection, rate optimization, and daytime management reduced the manual workload, so teams could focus on strategic planning and creative execution.

The broader takeaway

When you integrate machine learning into traditional media channels, you can achieve real-time adaptability and predictive performance typically reserved for digital platforms. By building SmartSpot360 on AWS, BMG360 was able to scale media buying operations efficiently, leveraging cloud-based data processing, machine learning models, and real-time analytics.

Key shifts enabled by AI and AWS infrastructure:

• Dynamic budget optimization based on real-time campaign data
• Predictive modeling for spot performance and audience trends
• Reduced time spent on manual media buying processes
• Improved scalability and faster adaptation to market changes
• Stronger integration between traditional and digital media strategies

Machine learning improves media buying efficiency and reshapes the way you plan, execute, and scale traditional advertising in a digital-first world.

Evolving SmartSpot360 with AWS

As SmartSpot360 continues to evolve, BMG360 is expanding the platform’s capabilities beyond media buying optimization to support broader planning and creative decision-making across channels.

Future development is focused on:

• Enhancing predictive models using reinforcement learning for deeper campaign intelligence
• Extending optimization to creative variants and message testing, allowing for more efficient creative to be identified and scaled faster than traditional methodologies
• Building a self-service planning interface for in-house teams and agency partners

Conclusion

When you run high-volume campaigns across TV and radio, the ability to optimize in real time, predict outcomes, and reduce manual overhead becomes a competitive advantage for both clients and agencies alike. SmartSpot360, built on AWS, gives you that edge, helping you scale performance, not just media spend.

As you continue to refine your media mix, the infrastructure you choose matters. Cloud-focused services, tools and machine learning capabilities give you the flexibility to evolve your planning, creative testing, and cross-channel coordination without sacrificing control or insight.

Traditional media doesn’t have to operate in a silo or fall behind digital channels. With a machine learning-powered solution you can bring precision, adaptability, and intelligence to every aspect of your media buying strategy.

BMG360 continues to evolve SmartSpot360 by leveraging the power of AWS, so they don’t just keep up with market shifts, they adapt in real time, align resources more effectively, and unlock greater value from every campaign.

To learn more about SmartSpot360 contact BMG360. Or contact an AWS Representative to know how we can help accelerate your business.

Further reading

• AWS for Advertising and Marketing
• Get started with API Gateway
• Building Generative AI into Marketing Strategies: A Primer
• Best practices and design patterns for building machine learning workflows with Amazon SageMaker Pipelines
• Automate model retraining with Amazon SageMaker Pipelines when drift is detected

About BMG360

BMG360 is a full-service performance marketing and lead generation company that helps businesses grow through proprietary technology, deep media expertise, and high-converting creative.