A Q&A with Orange 142 Paid Social Expert, Rich Lozano
By Calvin Scharffs
AI is rewriting the rules of social media advertising. Nothing, it seems, is unaffected. The way we target audiences is shifting, as are the creatives that we show them. And optimization is less about manual tweaks and more about guiding machine-driven systems. For brands that rely on social and social commerce as core revenue drivers, the stakes couldn’t be higher. Failing to adapt quickly and efficiently can spell the difference between owning a category and falling behind.
To understand what all of this means in practical terms, I spoke with Rich Lozano, Digital Advertising Manager at Orange 142. Rich spends his days deep in paid social, helping clients navigate the tension between automation and strategy, and finding ways to keep campaigns both effective and brand-safe in an AI-powered landscape.
CS: How is AI changing the way advertisers approach audience targeting on social platforms and in paid social?
RL: AI is reshaping paid social in a couple of ways. Platforms like Facebook have always relied on machine learning to optimize delivery—deciding which creatives to show to which audiences, at what times, based on performance signals. But it has evolved significantly. The optimization signals today are much stronger and can identify patterns that the systems couldn’t recognize years ago.
Each major platform now has its own branded AI optimization product. Meta is Advantage +, TikTok is called Smart+ Targeting, and Google has Performance Max. These tools enable advertisers to begin with minimal audience assumptions and let the AI refine the targeting automatically. The role of the advertiser then becomes deciding when to use broad targeting, when to lean into lookalikes or interest-based audiences, and how much to guide the AI versus letting it run.
Beyond targeting, AI tools are helping with creative versioning. Meta, for instance, will generate multiple ad copy variations from a single prompt, adding elements such as emojis or alternate calls-to-action. Externally, tools such as ChatGPT or Gemini can be used to expand one strong piece of copy into multiple testable variations. So rather than one or two options, advertisers can quickly create a half dozen versions to A/B test and optimize for best performance.
CS: What impact is AI having on creative strategy, especially when it comes to asset development and versioning for social media?
RL: In an ideal world, clients or agencies would have a creative team constantly producing new assets every few weeks, building on what’s worked and avoiding creative fatigue. The reality is, most teams don’t have that capacity. AI tools help fill the gap by quickly producing new copy based on what has already been successful, giving us momentum without requiring us to start from scratch.
For images and video, AI is continually improving, but its progress remains uneven. Meta, for example, can auto-generate images, but they rarely look professional enough to justify ad spend. Where AI does help is with simpler needs like B-roll or filler video that can support voice-overs and meet placement requirements. These options give overextended clients a way to keep fresh creative in rotation, even if it’s not perfect, and help campaigns avoid stagnation.
CS: How has campaign optimization evolved now that platforms are relying more on AI-driven delivery systems?
RL: Campaign optimization has improved significantly with AI-driven delivery. The key is to provide the system with strong, quality signals. At Orange 142, we use a conversion-first approach, such as requests for information (RFIs) for higher education or purchases for e-commerce. To achieve these outcomes, we must ensure that a robust data source is feeding the algorithm. Once those signals are in place, the system can optimize effectively. Over the past decade, as machine learning and AI have advanced, campaign performance has advanced alongside them.
CS: Are we seeing a shift in how performance is measured or understood in AI-powered campaigns?
RL: AI is changing how we think about performance. Too often, discussions focus only on upcoming campaigns. Less attention is given to current campaigns or past results. AI can help here. It can process data from previous quarters, uncover patterns, and highlight insights that humans might miss. There are simply too many signals for one person to review.
By applying AI to raw campaign data in a secure, privacy-compliant environment, we can generate insights at scale. These findings inform media buyers and account managers in making better decisions. Importantly, client data remains protected throughout the process.
The insights also help shape broader marketing strategies beyond a single campaign. These findings inform media buyers and account managers in making more informed decisions. They also help clients shape broader marketing strategies beyond a single campaign.
CS: What is the right balance between letting the platform optimize and maintaining strategic control as an advertiser?
RL: There is always a balance between automation and human judgment. The technology is powerful, but it lacks the nuance that comes from honest conversations with clients. Platforms like Meta’s Advantage+ can run broad optimizations, but they might overlook essential audience signals. For example, a museum campaign may rely heavily on grandparents bringing their grandchildren, something the AI would not identify on its own.
By adding those human insights into the system, we guide the AI in the right direction. It is like giving a search dog a scent to follow before letting it run. The best performance is achieved by combining both approaches: human-driven signals and AI-driven optimization. The synthesis of human strategy and machine optimization creates the strongest results.
CS: What new capabilities or limitations does AI introduce when it comes to testing and learning on social media or paid social campaigns?
RL: AI brings new strengths and weaknesses to testing and learning. On the positive side, it removes barriers and speeds up the process. For example, AI can generate multiple versions of ad copy from a single prompt, saving time and enabling teams to test more ideas quickly. It also improves targeting and optimization, though much of that happens inside a black box.
The limitations are most apparent in visual content. AI-generated images from platforms like Meta often look unpolished and can be brand-unsafe. Despite this, platforms are pushing hard for advertisers to adopt them, sometimes even enabling these features by default. We have seen clients run into embarrassing outcomes, such as distorted images, because the system generated poor creatives. When we set up campaigns, we turn off that feature.
In short, AI is a valuable tool for copy and targeting, but its visual outputs still need oversight. This is why agencies should resist the pressure to adopt half-ready features and focus on protecting client performance and brand integrity.
CS: Speaking of agencies, how is the role of the agency evolving in AI-driven paid social campaigns?
RL: It’s definitely shifting as AI handles more of the technical work. In the past, marketers relied heavily on specialized know-how. Today, tools automate much of that work, which allows agencies to focus on higher-level strategy and creative problem-solving.
However, agencies must now wear many hats. Basic coding or technical skills are useful, but what matters most is combining them with marketing fundamentals and strategic thinking. Clients are more informed than ever, so agencies must serve as thought leaders. That means helping clients understand what is coming next, how to use their budgets wisely, and how to plan for contingencies.
That begins with understanding how each platform uses AI, and the subtle nuances involved. It also requires looking closely at each platform to recognize which features are right for each brand. Orange 142 helps brands and agencies navigate the complexities of AI-powered social platforms with hands-on support and tailored strategies. We don’t just launch campaigns; we manage every detail, from audience strategy to content execution and performance optimization.
Our approach blends real human insight with platform-level automation, ensuring your paid social campaigns stay true to your brand and deliver measurable results.
CS: Do you have any parting thoughts for our readers?
RL: Yes. There is an idea I keep coming back to: almost anything can be accomplished if we are willing to think creatively and utilize the tools available. It might sound simple, but adopting a “we can do it” mindset changes how we approach challenges.
For example, if a client wants to re-engage customers who haven’t purchased in 28 days, we can set up an automated text to reach them at a specific time of day. Tools already exist to make that happen, and often clients are surprised by how obvious or straightforward these solutions are once implemented.
Too often, capabilities are left untapped. If we think beyond the standard playbook and strive for excellence, we can uncover ideas that impress clients and strengthen our partnerships. The combination of creativity, ambition, and available technology allows us to deliver solutions that may not have been considered before.
Want to learn how this can work for you? Orange 142 helps SMBs navigate and maximize emerging advertising channels with strategic guidance and best practices. Let’s connect to explore the right approach for your goals.
