In a recent PPC Live podcast conversation, Dean Kadi, Head of Paid Growth at One Link Media, shared a real-world agency experience where a client insisted on replacing high-performing Meta ads with heavily branded creative — despite clear evidence that the existing strategy was delivering strong results. The discussion highlights the tension agencies often face between expertise and client preference, while offering valuable lessons on communication, testing, tracking, and why data should always lead decision-making in PPC.
The campaign was performing exceptionally well
Dean Kadi and his team at One Link Media built a highly successful Meta advertising strategy for premium woodworking brand Rubio Monocoat using user-generated content (UGC). By testing multiple creators, hooks, formats, and messaging angles, they improved the account’s ROAS from around 2.1x to consistently between 3x and 4x. Their testing revealed that the biggest purchase driver was not the product’s variety of colours, but the fact that customers only needed one coat of product, saving significant time and effort.
The client wanted to pause all winning ads
Despite strong performance, the client unexpectedly requested that all successful UGC ads be paused in favour of heavily branded static and video creatives. The new ads looked polished but failed to feel native to the Meta platform, which is often critical for engagement and conversion. The decision wasn’t based on performance issues but on the client’s preference for more traditional branding.
The dangerous assumption behind the new strategy
The client based their new creative direction on a survey suggesting customers liked the brand’s colour range, assuming this was the primary reason people purchased. However, the agency’s testing data had already proven otherwise. This highlights a common mistake in marketing where internal assumptions or isolated feedback override broader performance data and real-world customer behaviour.
“We’d prefer this to be a winner”
One of the most telling moments in the discussion came when the client admitted they simply wanted the new creative approach to succeed. Dean pointed out that paid media doesn’t work based on preference or hope — audiences decide what resonates. No matter how strongly stakeholders feel about a campaign direction, performance data ultimately determines success.
What agencies should do in situations like this
Dean advised agencies to stay calm, professional, and evidence-led when disagreements with clients arise. Rather than arguing emotionally, marketers should clearly communicate risks, explain their reasoning, and document recommendations in writing. By maintaining professionalism and allowing the data to speak for itself, agencies can protect relationships while still standing behind their expertise.
The results tanked — exactly as expected
The new branded creatives quickly underperformed, with rising acquisition costs and declining efficiency across Meta campaigns. Although the agency continued testing audiences and optimisation strategies, the core issue remained the creative itself. After approximately eight weeks of poor results, it became clear that the client’s new direction was not working.
Returning to UGC restored performance
Once the client agreed to reintroduce the original UGC ads, campaign performance improved rapidly within just a couple of weeks. The return of native-looking content and proven messaging angles restored the account’s efficiency and validated the agency’s original strategy. Interestingly, Google Ads performance remained relatively stable because those campaigns relied more heavily on branded search activity.
The bigger lesson: let data tell the story
Dean’s biggest takeaway was that agencies should rely on data rather than emotion when navigating difficult client situations. Sometimes clients need to see underperformance firsthand before accepting recommendations. By consistently presenting clear reporting and measurable outcomes, marketers can use evidence to guide conversations and rebuild trust.
Common PPC mistakes agencies still encounter
Beyond this client story, Dean highlighted poor tracking setup as one of the most common mistakes still seen in PPC accounts today. Missing server-side tracking, incorrect event configurations, and weak conversion tracking setups can severely impact optimisation and reporting. Even the strongest campaigns struggle if the underlying data infrastructure is flawed.
AI won’t fix a bad strategy
Dean also warned against overreliance on AI tools in marketing. While AI can improve efficiency and speed up workflows, it cannot compensate for weak strategy or poor thinking. Marketers still need to critically evaluate outputs, refine prompts, and apply human judgment, because clients ultimately hold people accountable — not AI systems.
Final thoughts
This story serves as a reminder that successful PPC campaigns rely on testing, data, and strategic discipline rather than internal opinions or branding preferences alone. Agencies must balance professionalism with confidence in their expertise, document their recommendations carefully, and trust performance metrics over assumptions. In the end, audiences decide what works — and the data almost always reveals the truth.
Search Engine Land is owned by Semrush. We remain committed to providing high-quality coverage of marketing topics. Unless otherwise noted, this page’s content was written by either an employee or a paid contractor of Semrush Inc.
