We are currently living through an unmitigated gold rush of automated efficiency. Every morning our LinkedIn feeds are flooded with new tech announcements and talk of AI agents that promise to transform the tedious, soul-crushing administrative slog of performance marketing into a cool summer breeze. But in my day-to-day work observing high-spending acquisition teams, it’s very rarely this straightforward.
More often than not, I find that marketers, founders, and user acquisition (UA) professionals are trying to use AI and automation as structural band-aids. They are trying to build complex, technical engineering solutions to patch up deep, fundamental holes in their operational strategy, their human communication, and their data hygiene.
For this article I interviewed Filippo de Rose, Chief Marketing Officer at the self-care coaching app, Fabulous.
My goal was to understand how a massive, high-performing performance marketing operation survives the modern transition into an AI-driven ecosystem without sacrificing its strategic focus, its budget, or its culture.
Filippo is responsible for high-volume ad spend across global ad networks. But what makes his perspective so valuable isn't a single, spectacular scaling metric or one-off campaign success. It's the fact that he has spent the last few years quietly, methodically executing a masterclass in building a lean, semi-automated acquisition engine.
Here we learn from Fabulous’ real-world processes and Filipo’s marketing philosophies on how to approach automation not as a defensive shortcut, but as a strategic amplifier.
Key takeaways
The “software first” hiring framework. Stop viewing headcount as your default operational lever. Treat automation as a potential alternative for repetitive administrative tasks, freeing human capital for creative and strategic leverage.
Automation protects peace of mind. True ROI isn’t always captured on an isolated, backward-looking spreadsheet. The real economic value of a robust automation framework lies in risk mitigation, error reduction, and freeing up creative mental capacity.
Data hygiene is your AI bedrock. If your accounts lack naming taxonomy, structured testing frameworks, and pristine data pipelines, any automation tool you connect will simply scale your structural mess.
The changing UA skillset. The modern media buyer’s role is shifting away from granular level bidding controls toward defining high-level strategic inputs, testing hypotheses, and setting strict guardrails for algorithmic platforms.
The slow burn: moving past the myth of the success “trigger”
More often than not, the automation related case studies and social posts that fill our feeds sell a dramatic, cinematic narrative. We are told about the client who exponentially scaled their ads overnight with “one simple prompt,” the catastrophic regional expansion that went wrong, or the frantic tracking crisis that forced a team to tear down their manual structures and rebuild from scratch (with a team of Claude agents nonetheless). But real organizational transformation rarely happens because of a sudden event. It often occurs as a slow burn. It’s a quiet realization that the ground under the industry has shifted, that the tools have democratized, and that continuing to run campaigns manually isn’t just inefficient, it’s an active operational risk.
When Filippo and his team at Fabulous began auditing their media buying workflows a few years ago, they didn't face a sudden administrative emergency. They faced the universal reality of the modern advertising ecosystem: the sheer volume of creative variants, audience tests, and algorithmic updates required to remain competitive was expanding, while human hours remained strictly finite.
For us, moving toward automation wasn’t driven by an overnight emergency or a dramatic budget spike. It was a gradual, highly intentional recognition of changing industry dynamics. The platforms were opening up their APIs, machine learning was becoming table stakes, and we realized that keeping our media buyers tethered to manual monitoring was keeping them away from the work that actually moves revenue.
Filippo De Rose, CMO at Fabulous
This psychological shift completely reframed how Fabulous looks at talent acquisition and team structures. In a traditional performance team, when production demand goes up, the immediate management instinct is often to increase headcount. Throw more human bodies to fulfill the workload. But if you look closely at the unit economics of a mid-sized growth team, adding headcount before optimizing your technical foundations is a surefire way to generate bureaucracy and decrease your agility.
Filippo flipped the script on this tradition entirely. He implemented a deeply ingrained structural rule: When a new task, a repetitive workflow, or an operational gap emerges within the UA team, the first instinct must be to design an automated or software-driven solution before considering a new hire.
Every time we identify a recurring operational need, my first question is never 'Who should we hire for this?' It’s always: ‘Can this be solved by an API, an internal script, or an automated system?’ We treat software as our baseline operational layer. We only look to expand our human headcount when a problem requires deep, strategic critical thinking, emotional intelligence, or qualitative creative ideation that no machine can replicate.
The lesson here is that automation scales robust, well-vetted processes just as efficiently as it accelerates broken ones. If your account governance is a mess, your data inputs are unrefined, and your team isn’t aligned, adding an advanced automation engine will only allow you to make mistakes faster than ever before.
Moving from broad intuition to rules-driven governance
To understand how this philosophy translates into day-to-day media buying, we can review the practical, programmatic logic running under the hood of Fabulous’s ad accounts. They don't rely on broad media buyer intuition or ad-hoc manual checks. Instead, they have turned their operational memory into a series of automated guardrails that execute around the clock.
Here are the three core automation use cases that Filippo relies on to maintain a lean, high-performing user acquisition framework.
Use case 1: curbing weekend creative fatigue
One of the most clear-cut battles every performance marketer fights is the reality of weekend creative decay. You spend all week designing, validating, and launching a fresh batch of assets. By Friday afternoon, a handful of variations are showing incredible promise. But then, the weekend. Consumer behavior shifts, delivery algorithms recalibrate, and by Sunday night, those winning assets are demonstrating creative fatigue. Those fatigued ads will happily burn through ad budgets before anyone logs back in on Monday morning.
Filippo’s team automated this entire protective workflow. They set up a programmatic rule that continuously tracks early behavioral and performance indicators, specifically focusing on metrics like a sudden drop in Click-Through Rate (CTR) alongside an increase in ad frequency.
We built clear automated parameters to manage our creative testing over the weekend. If an active creative asset experiences a dramatic performance dip or hits a specific fatigue threshold when the team is away, the system pauses it instantly. We no longer return to the office on Monday morning to find out we spent thousands of dollars on an asset that stopped converting 48 hours prior.
Use case 2: The systematic tracking of high-performance signals
Automation isn't merely a defensive mechanism designed to prevent financial loss, it is an offensive growth accelerator. The classic challenge in manual media buying is that when an ad asset begins to exhibit breakthrough performance, human media buyers are naturally risk-averse. They hesitate. They worry about resetting the ad platform's learning phase, or they wait for days of data confirmation before adjusting budgets. Until they decide to act, the competitive window has often narrowed.
Fabulous’ engine relies on automated scaling rules that look for highly specific conversion signals. When an ad set or a specific creative variation clears its baseline CPA thresholds over a validated statistical sample, the automation engine systematically bumps the budget by an optimized percentage, maximizing its reach while. This allows the team more time to focus on creative strategy.
When we find a legitimate creative winner, we don't want to wait around for a manual review. Our scaling rules are built to detect high-performance signals in real-time. If an ad variant is operating cleanly below our target cost-per-action, the system automatically feeds it more capital within our preset strategic limits.
Use case 3: automating ad naming
One of the biggest challenges in being able to truly take advantage of AI and automation in performance marketing is overcoming the lack of standardized datasets. If one media buyer names a campaign US_M_Pro_FB, another names it USA_Male_Promo_1025, and a third completely forgets to include the localization code, your tracking environment fragments instantly. You quickly find yourself stuck in a classic “Spreadsheet of Doom,” wasting hours of expensive engineering time manually stitching rows together in Excel just to find a single source of truth.
Filippo recognized that to build a sustainable, automation-ready infrastructure, data hygiene had to be non-negotiable. Fabulous implemented an automated naming convention governance system. If a campaign, ad set, or asset is launched without adhering to the precise, structural nomenclature required by their data ecosystem, their automation engine flags it immediately, maintaining a pristine database that can feed clean information back into their predictive reporting tools.
You cannot build a sophisticated automated framework on top of sloppy data. We enforce an absolute, rigid naming taxonomy across every single channel. If your tracking codes, nomenclature, or structural tags are inconsistent, your automation rules won't work, and your data analysis becomes completely useless. Squeaky clean inputs are the prerequisite for automated scale.
The invisible ROI: the economic reality of “Peace of Mind”
One persistent question in the world of automation and AI is ‘how much time is being saved?’ We are living in a pandemic of perceived productivity. Senior stakeholders want to know exactly how many human hours these systems result in. What is the precise, immediate reduction in our operational overhead.
But as anyone who has actually managed high-stakes acquisition campaigns knows, the most profound, transformative returns on automation are completely invisible on a standard, backward-looking spreadsheet.
You cannot easily put a price tag on peace of mind. You cannot draw a neat, simple graph that captures the exact dollar value of a media buying team that is no longer experiencing chronic burnout, emotional exhaustion, or the constant low-grade anxiety of manually babysitting accounts over weekends and holidays.
Funnily enough, I have never treated automation as an isolated metric designed to slash human hours on a sheet. I look at it from a structural standpoint. What is the financial cost of a human media buyer who makes an administrative mistake on a Saturday night because they are tired? What is the cost of our top strategic talent quitting because they spend 80% of their week copy-pasting creatives instead of thinking? Automation gives our team the cognitive breathing room to do their best work.
When you remove the repetitive administrative weight of performance marketing, the manual tracking adjustments, the formatting checks, the baseline budget shifts, you aren't simply saving time, you are investing in creative focus. You are freeing up your growth leads to move away from reactive troubleshooting and toward proactive, high-level hypothesis testing.
Instead of spending their mornings reacting to yesterday's metrics, Filippo’s team can dedicate their mental energy to deep strategic exploration:
Which unique audience personas have we completely failed to communicate with?
What macro-economic trends or industry changes are altering our target market's psychological relationship with health apps?
How can we structure our creative production to build true, long-term brand differentiation rather than endlessly copying industry trends?
Automation didn't minimize my role or reduce our team's value. It elevated it. It forced us to grow from platform executioners into high-level strategic thinkers. It allowed us to carefully evaluate our budget allocation, manage massive global spend with an incredibly lean footprint, and ensure that every pound we invest is protected by clear, rule-based governance.
The team of the future: relishing the era of strategic constraints
We are rapidly exiting the era of the platform specialist. For the last decade, performance marketing heavily rewarded the individuals who knew how to manually navigate the technical complexities of ad platforms. The people who mastered manual bidding, precise lookalike stacking, and complex dayparting.
But as ad platforms continue to evolve into highly automated ecosystems, those manual, transactional skills are becoming less essential. The native algorithms are taking total control over granular optimization, forcing media buyers to transform or get left behind.
The entire definition of a performance marketing role has transformed. If your entire career is anchored in manual optimization, the algorithms are going to outpace you. The real winners of this new era are the growth marketers who understand data architecture, who know how to speak to creative teams in an objective, analytical language, and who can build a unified operational system that runs flawlessly without needing constant human intervention.
So, how do marketing teams successfully create and scale user acquisition campaigns in this new era of AI and automation? According to Filippo, it requires a strict focus on communication, commitment to governance and a culture that prioritizes creative strategy.
We are currently living through an unmitigated gold rush of automated efficiency. Every morning our LinkedIn feeds are flooded with new tech announcements and talk of AI agents that promise to transform the tedious, soul-crushing administrative slog of performance marketing into a cool summer breeze. But in my day-to-day work observing high-spending acquisition teams, it’s very rarely this straightforward.
More often than not, I find that marketers, founders, and user acquisition (UA) professionals are trying to use AI and automation as structural band-aids. They are trying to build complex, technical engineering solutions to patch up deep, fundamental holes in their operational strategy, their human communication, and their data hygiene.
For this article I interviewed Filippo de Rose, Chief Marketing Officer at the self-care coaching app, Fabulous.
My goal was to understand how a massive, high-performing performance marketing operation survives the modern transition into an AI-driven ecosystem without sacrificing its strategic focus, its budget, or its culture.
Filippo is responsible for high-volume ad spend across global ad networks. But what makes his perspective so valuable isn't a single, spectacular scaling metric or one-off campaign success. It's the fact that he has spent the last few years quietly, methodically executing a masterclass in building a lean, semi-automated acquisition engine.
Here we learn from Fabulous’ real-world processes and Filipo’s marketing philosophies on how to approach automation not as a defensive shortcut, but as a strategic amplifier.
Key takeaways
The “software first” hiring framework. Stop viewing headcount as your default operational lever. Treat automation as a potential alternative for repetitive administrative tasks, freeing human capital for creative and strategic leverage.
Automation protects peace of mind. True ROI isn’t always captured on an isolated, backward-looking spreadsheet. The real economic value of a robust automation framework lies in risk mitigation, error reduction, and freeing up creative mental capacity.
Data hygiene is your AI bedrock. If your accounts lack naming taxonomy, structured testing frameworks, and pristine data pipelines, any automation tool you connect will simply scale your structural mess.
The changing UA skillset. The modern media buyer’s role is shifting away from granular level bidding controls toward defining high-level strategic inputs, testing hypotheses, and setting strict guardrails for algorithmic platforms.
The slow burn: moving past the myth of the success “trigger”
More often than not, the automation related case studies and social posts that fill our feeds sell a dramatic, cinematic narrative. We are told about the client who exponentially scaled their ads overnight with “one simple prompt,” the catastrophic regional expansion that went wrong, or the frantic tracking crisis that forced a team to tear down their manual structures and rebuild from scratch (with a team of Claude agents nonetheless). But real organizational transformation rarely happens because of a sudden event. It often occurs as a slow burn. It’s a quiet realization that the ground under the industry has shifted, that the tools have democratized, and that continuing to run campaigns manually isn’t just inefficient, it’s an active operational risk.
When Filippo and his team at Fabulous began auditing their media buying workflows a few years ago, they didn't face a sudden administrative emergency. They faced the universal reality of the modern advertising ecosystem: the sheer volume of creative variants, audience tests, and algorithmic updates required to remain competitive was expanding, while human hours remained strictly finite.
For us, moving toward automation wasn’t driven by an overnight emergency or a dramatic budget spike. It was a gradual, highly intentional recognition of changing industry dynamics. The platforms were opening up their APIs, machine learning was becoming table stakes, and we realized that keeping our media buyers tethered to manual monitoring was keeping them away from the work that actually moves revenue.
Filippo De Rose, CMO at Fabulous
This psychological shift completely reframed how Fabulous looks at talent acquisition and team structures. In a traditional performance team, when production demand goes up, the immediate management instinct is often to increase headcount. Throw more human bodies to fulfill the workload. But if you look closely at the unit economics of a mid-sized growth team, adding headcount before optimizing your technical foundations is a surefire way to generate bureaucracy and decrease your agility.
Filippo flipped the script on this tradition entirely. He implemented a deeply ingrained structural rule: When a new task, a repetitive workflow, or an operational gap emerges within the UA team, the first instinct must be to design an automated or software-driven solution before considering a new hire.
Every time we identify a recurring operational need, my first question is never 'Who should we hire for this?' It’s always: ‘Can this be solved by an API, an internal script, or an automated system?’ We treat software as our baseline operational layer. We only look to expand our human headcount when a problem requires deep, strategic critical thinking, emotional intelligence, or qualitative creative ideation that no machine can replicate.
The lesson here is that automation scales robust, well-vetted processes just as efficiently as it accelerates broken ones. If your account governance is a mess, your data inputs are unrefined, and your team isn’t aligned, adding an advanced automation engine will only allow you to make mistakes faster than ever before.
Moving from broad intuition to rules-driven governance
To understand how this philosophy translates into day-to-day media buying, we can review the practical, programmatic logic running under the hood of Fabulous’s ad accounts. They don't rely on broad media buyer intuition or ad-hoc manual checks. Instead, they have turned their operational memory into a series of automated guardrails that execute around the clock.
Here are the three core automation use cases that Filippo relies on to maintain a lean, high-performing user acquisition framework.
Use case 1: curbing weekend creative fatigue
One of the most clear-cut battles every performance marketer fights is the reality of weekend creative decay. You spend all week designing, validating, and launching a fresh batch of assets. By Friday afternoon, a handful of variations are showing incredible promise. But then, the weekend. Consumer behavior shifts, delivery algorithms recalibrate, and by Sunday night, those winning assets are demonstrating creative fatigue. Those fatigued ads will happily burn through ad budgets before anyone logs back in on Monday morning.
Filippo’s team automated this entire protective workflow. They set up a programmatic rule that continuously tracks early behavioral and performance indicators, specifically focusing on metrics like a sudden drop in Click-Through Rate (CTR) alongside an increase in ad frequency.
We built clear automated parameters to manage our creative testing over the weekend. If an active creative asset experiences a dramatic performance dip or hits a specific fatigue threshold when the team is away, the system pauses it instantly. We no longer return to the office on Monday morning to find out we spent thousands of dollars on an asset that stopped converting 48 hours prior.
Use case 2: The systematic tracking of high-performance signals
Automation isn't merely a defensive mechanism designed to prevent financial loss, it is an offensive growth accelerator. The classic challenge in manual media buying is that when an ad asset begins to exhibit breakthrough performance, human media buyers are naturally risk-averse. They hesitate. They worry about resetting the ad platform's learning phase, or they wait for days of data confirmation before adjusting budgets. Until they decide to act, the competitive window has often narrowed.
Fabulous’ engine relies on automated scaling rules that look for highly specific conversion signals. When an ad set or a specific creative variation clears its baseline CPA thresholds over a validated statistical sample, the automation engine systematically bumps the budget by an optimized percentage, maximizing its reach while. This allows the team more time to focus on creative strategy.
When we find a legitimate creative winner, we don't want to wait around for a manual review. Our scaling rules are built to detect high-performance signals in real-time. If an ad variant is operating cleanly below our target cost-per-action, the system automatically feeds it more capital within our preset strategic limits.
Use case 3: automating ad naming
One of the biggest challenges in being able to truly take advantage of AI and automation in performance marketing is overcoming the lack of standardized datasets. If one media buyer names a campaign US_M_Pro_FB, another names it USA_Male_Promo_1025, and a third completely forgets to include the localization code, your tracking environment fragments instantly. You quickly find yourself stuck in a classic “Spreadsheet of Doom,” wasting hours of expensive engineering time manually stitching rows together in Excel just to find a single source of truth.
Filippo recognized that to build a sustainable, automation-ready infrastructure, data hygiene had to be non-negotiable. Fabulous implemented an automated naming convention governance system. If a campaign, ad set, or asset is launched without adhering to the precise, structural nomenclature required by their data ecosystem, their automation engine flags it immediately, maintaining a pristine database that can feed clean information back into their predictive reporting tools.
You cannot build a sophisticated automated framework on top of sloppy data. We enforce an absolute, rigid naming taxonomy across every single channel. If your tracking codes, nomenclature, or structural tags are inconsistent, your automation rules won't work, and your data analysis becomes completely useless. Squeaky clean inputs are the prerequisite for automated scale.
The invisible ROI: the economic reality of “Peace of Mind”
One persistent question in the world of automation and AI is ‘how much time is being saved?’ We are living in a pandemic of perceived productivity. Senior stakeholders want to know exactly how many human hours these systems result in. What is the precise, immediate reduction in our operational overhead.
But as anyone who has actually managed high-stakes acquisition campaigns knows, the most profound, transformative returns on automation are completely invisible on a standard, backward-looking spreadsheet.
You cannot easily put a price tag on peace of mind. You cannot draw a neat, simple graph that captures the exact dollar value of a media buying team that is no longer experiencing chronic burnout, emotional exhaustion, or the constant low-grade anxiety of manually babysitting accounts over weekends and holidays.
Funnily enough, I have never treated automation as an isolated metric designed to slash human hours on a sheet. I look at it from a structural standpoint. What is the financial cost of a human media buyer who makes an administrative mistake on a Saturday night because they are tired? What is the cost of our top strategic talent quitting because they spend 80% of their week copy-pasting creatives instead of thinking? Automation gives our team the cognitive breathing room to do their best work.
When you remove the repetitive administrative weight of performance marketing, the manual tracking adjustments, the formatting checks, the baseline budget shifts, you aren't simply saving time, you are investing in creative focus. You are freeing up your growth leads to move away from reactive troubleshooting and toward proactive, high-level hypothesis testing.
Instead of spending their mornings reacting to yesterday's metrics, Filippo’s team can dedicate their mental energy to deep strategic exploration:
Which unique audience personas have we completely failed to communicate with?
What macro-economic trends or industry changes are altering our target market's psychological relationship with health apps?
How can we structure our creative production to build true, long-term brand differentiation rather than endlessly copying industry trends?
Automation didn't minimize my role or reduce our team's value. It elevated it. It forced us to grow from platform executioners into high-level strategic thinkers. It allowed us to carefully evaluate our budget allocation, manage massive global spend with an incredibly lean footprint, and ensure that every pound we invest is protected by clear, rule-based governance.
The team of the future: relishing the era of strategic constraints
We are rapidly exiting the era of the platform specialist. For the last decade, performance marketing heavily rewarded the individuals who knew how to manually navigate the technical complexities of ad platforms. The people who mastered manual bidding, precise lookalike stacking, and complex dayparting.
But as ad platforms continue to evolve into highly automated ecosystems, those manual, transactional skills are becoming less essential. The native algorithms are taking total control over granular optimization, forcing media buyers to transform or get left behind.
The entire definition of a performance marketing role has transformed. If your entire career is anchored in manual optimization, the algorithms are going to outpace you. The real winners of this new era are the growth marketers who understand data architecture, who know how to speak to creative teams in an objective, analytical language, and who can build a unified operational system that runs flawlessly without needing constant human intervention.
So, how do marketing teams successfully create and scale user acquisition campaigns in this new era of AI and automation? According to Filippo, it requires a strict focus on communication, commitment to governance and a culture that prioritizes creative strategy.
Scott is an experienced marketer, copywriter, podcaster and consultant with expertise spanning everything from performance marketing and UX to data analytics and AI strategy. He loves exploring the intricacies of marketing and turning complex, technical subjects into user friendly content.
