A seasoned mountaineer once shared an observation that stayed with me. He said the mountain never defeats people because it changes. It defeats them because climbers keep trusting maps drawn from yesterday’s weather. So what I am writing is quite opinionated and some leaders may not agree with me, and that’s okay. This is a piece I am writing to debate, so feel free to comment.
In my observation, most Brands Are Still Optimizing for a World That No Longer Exists.
Performance marketing has entered a similar phase.
For years, digital marketing rewarded those who could master platforms, ‘manipulate’ bidding strategies, identify the right audience segments, and continuously optimize campaigns. There was comfort in precision. Agencies proudly showcased complex audience pyramids, hundreds of micro-segments, and endless A/B tests. Brands became increasingly dependent on dashboards that appeared to explain everything.
Today, many of those maps seem to have lost their reliability.
The discipline itself has not become less scientific. It has become more probabilistic. AI has fundamentally altered (and still altering further) how campaigns are delivered, how audiences are identified, and how platforms interpret intent. Performance marketing is no longer just about finding customers. Increasingly, it is about helping algorithms discover the right customers for you.
That subtle shift has changed everything. So let me share few points.
From Audience Buying to Signal Engineering
Around 10 years ago, campaign success depended heavily on how well marketers defined audiences. Lookalike audiences, detailed interest targeting, demographic layering, exclusion logic, and keyword segmentation formed the foundation of digital campaigns. Agencies invested enormous effort into identifying who the customer was before platforms could find them.The marketer controlled the machine.
Today, platforms increasingly expect the opposite.
Meta’s Advantage + ecosystem, Google’s AI-powered Performance Max campaigns, broad targeting models, automated bidding, predictive attribution, and machine learning optimization all indicate one clear direction. Platforms want marketers to provide stronger signals rather than narrower instructions.
The competitive advantage has moved from audience selection towards data quality.
First-party customer data, high-quality conversion signals, creative diversity, server-side tracking, and clean event architecture now influence campaign performance far more than building increasingly sophisticated audience trees. Ironically, many organizations continue investing disproportionate effort in targeting while underinvesting in the quality of signals entering the algorithm.
3 Challenges that I observe Performance Marketing Is Grappling With
The first challenge is what I call the signal deficit.
Privacy regulations, Apple’s App Tracking Transparency framework, browser cookie restrictions, consent management requirements, and ongoing platform algorithm updates have significantly reduced observable customer behaviour.
Campaign managers often complain that attribution has become inaccurate. The larger issue is not attribution alone. Platforms are learning from fewer reliable behavioural signals than before.Many organizations continue making optimization decisions using incomplete data while assuming the data remains complete.
Forward-looking agencies are responding by implementing server-side tracking, strengthening CRM integrations, adopting Conversions API frameworks, improving event prioritisation, and investing heavily in first-party data infrastructure. The objective is no longer collecting more data. It is collecting more trustworthy data.
The second challenge is creative fatigue happening faster than algorithmic learning.
AI has lowered production costs dramatically. Brands now produce more creatives than ever before. Unfortunately, quantity has not translated into effectiveness.Consumers are exposed to thousands of commercial messages daily. Algorithms quickly identify declining engagement, making creative deterioration significantly faster than it was a decade ago.
Performance teams are seeming to still treat creative as an asset delivered by another department instead of treating it as the primary optimization variable.
The agencies producing stronger outcomes have reorganized themselves differently. Performance specialists now collaborate with creative strategists from the beginning. Data influences storytelling, and storytelling generates new data. Creative testing has become a weekly operating rhythm rather than a quarterly campaign exercise.
The third challenge is fragmented measurement.
Most brands continue reviewing Meta, Google, Amazon, CRM, website analytics, marketplaces, and offline sales as separate reporting ecosystems. When measurement remains fragmented, optimization becomes fragmented too. I observe that global agencies are now building unified measurement frameworks that combine incrementality testing, media mix modelling, customer lifetime value, assisted conversions, and business profitability rather than platform-specific ROAS alone.
In my own words I quote that “Performance marketing is slowly becoming business optimization instead of media optimization”.
And,…..this distinction matters.
What Is Quietly Not Working
- One practice continues to concern me across both agencies and brands.
- Too many decisions are being made because dashboards recommend them.
- Algorithms are incredibly effective at recognizing behavioural patterns.
- They cannot independently understand strategic context.
- An algorithm may identify the cheapest acquisition opportunity. It cannot determine whether those customers strengthen long-term brand equity.
- Similarly, automated bidding can optimize towards today’s conversion objective. It cannot determine whether today’s optimization weakens tomorrow’s pricing power.
- Performance marketing has become extraordinarily automated.
- Judgement has become extraordinarily valuable.
- The agencies creating sustainable business growth are not those that rely least on artificial intelligence. They are the ones that know precisely where artificial intelligence should stop and human judgement should begin.
A Different Operating Model? (don’t kill me for this) as I understand client servicing more as ‘hospitality’ rather than just servicing.
Perhaps the biggest opportunity lies in dismantling one of the oldest structural problems inside agencies.
Creative teams produce advertisements.
Media teams distribute them.
Performance teams optimize them.
Analytics teams report them.
Customers never experience those departments separately.
High-performing agencies are increasingly replacing departmental ownership with integrated growth pods where strategists, creative thinkers, media planners, analysts, and CRM specialists work toward one commercial outcome instead of multiple departmental outputs.
Marketing performance becomes significantly stronger when optimization happens before campaigns launch rather than after reports are generated. Integration is becoming an operational advantage rather than merely a cultural aspiration.
The One Technical Capability for 2027
The single most valuable technical capability could be say…signal architecture.
Organizations that build robust first-party data ecosystems, clean conversion infrastructure, server-side tracking, identity resolution, predictive customer models, and AI-ready measurement frameworks will consistently outperform organizations chasing platform hacks. Algorithms will increasingly reward signal quality over manual optimization. Companies that own better data ecosystems will naturally build stronger advertising ecosystems.
The One Leadership Capability for 2027
The defining leadership capability will be…decision synthesis.
Marketing leaders may face unprecedented volumes of automated recommendations, predictive analytics, attribution models, AI-generated insights, and optimization suggestions. The challenge will not be accessing information. The challenge will be knowing which information deserves action. (this is nothing new, however, do we focus enough on this or just keep trapped inside the walled gardens without even trying to navigate)
Also, here I accentuate that exceptional leaders should connect commercial priorities, customer psychology, creative intuition, technological capability, and organizational judgment into one coherent decision system. They will ask whether the algorithm is solving the right business problem before asking whether the campaign is performing well. When we try asking that, even Brands I see are not interested in this conversation maybe because they don’t look at long term value.
. Businesses that combine robust technical infrastructure with disciplined strategic judgment will create durable competitive advantage
________________________________________________________________________________________________
The views and opinions published here belong to the author and do not necessarily reflect the views and opinions of the publisher.
