This is the second installment in our “Gut Check” series. Previously, Part-1 explored Multi-Surface Organic and the tectonic shifts reshaping how brands get discovered across AI-powered surfaces. In Part-3 later in February, we’ll tie it all together, examining how paid and organic are merging into unified brand discovery systems. For this installment, let’s dig into what’s happening in the paid media trenches.
TL;DR
Premise: Paid performance media in 2026 continues to evolve at a breakneck pace. AI-powered Search, Performance Max transparency upgrades, Bayesian incrementality testing, keywordless targeting, and last year’s tariff-driven margin compression have fundamentally altered how marketers approach our craft. Platforms are racing toward full automation while simultaneously (and finally) giving us back some visibility into the black boxes of current-day marketing. Practitioners who thrive will be those who understand this isn’t a zero-sum game between human expertise and machine learning, never was. Reality is all about teaching the machines what matters and knowing when to override them.
Expert Sources: Frederick Vallaeys (Optmyzr), Brad Geddes (Adalysis), Dennis Goedegebuure (Founder/Industry Strategist), Kirk Williams (ZATO), Aaron Levy (Optmyzr), Navah Hopkins (Microsoft), Greg Finn (Cypress North), Brooke Osmundson (Smith Micro Software), Sam Tomlinson (Warschawski), Ginny Marvin (Google), Jyll Saskin Gales (Google Ads Coach), Duane Brown (Take Some Risk), Casey Gill (Iterate.ai), Nils Rooijmans (Google Ads Performance Architect), Mike Ryan (smec), Ed Leake (God Tier Ads), Thomas Eccel (AdSea Innovations), AJ Wilcox (B2Linked), Julie Bacchini (Neptune Moon), Andrew Lolk (SavvyRevenue), Christine Zirnheld (Cypress North, Marketing O’Clock), Shawn Walker (BidWaves, Symphonic Digital), Wijnand Meijer (TrueClicks), and the broader paid search community who’ve been wrestling with these questions in real time.
Verdict: Generally convergent thinking. Remarkable level of consistency regarding direction and insightful unique takes on critical elements.
Shifts:
- Automation layering replaces automation resistance. The platforms will automate. Your job is to layer intelligence on top: guardrails, monitoring systems, and business logic the algorithms aren’t factoring.
- Incrementality becomes measurable at human budgets. Bayesian models and Google’s new testing frameworks make it possible to prove true lift with $5K and seven days, as opposed to $100K and six weeks.
- AI-powered Search signals a keywordless future. While keywords aren’t quite dead yet, the writing is on the wall. Intent-based matching will dominate, and the practitioners who understand this transition will have an edge.
- Performance Max transparency arrives, finally. Channel-level reporting, asset-level metrics, search term insights, and negative keywords at campaign level. The black box now has windows!
- First-party data becomes the decisive advantage. While many marketers focus on automation and signal loss, the brands with strong customer data and conversion infrastructure will outperform those relying on platform defaults.
- Economic volatility demands marketing-finance alignment. Last year’s bombardment of tariffs and ongoing margin compression mean PPC can no longer operate in a silo. CAC must be understood in context of COGS, LTV, and contribution margin.
- Account structure simplification is mandatory. Over-segmentation destroys signals. Consolidate until the data proves you need to separate.
- Demand gen emerges as the next battleground. Video Action Campaigns sunset, lookalike audiences, and full-funnel control make Demand Gen campaigns critical for 2026.
New Metrics:
- Incremental lift at realistic budgets (Bayesian posterior probability)
- True channel contribution (not platform-reported ROAS)
- Customer acquisition cost relative to contribution margin
- First-party data match rates across platforms
- Keywordless vs. keyword-driven query distribution
- Marginal ROAS (the ROAS of the last dollar spent, not the average)
- Clicked vs. Bought product discrepancy
Frederick Vallaeys: The Incrementality Imperative
Frederick Vallaeys, co-founder and CEO of Optmyzr and one of Google’s first 500 employees, has been leading the charge around measurement challenges for two decades. His work on incrementality testing cuts to the heart of what changed last year and what finally works.
In his December 2025 piece for Search Engine Land, How Bayesian testing lets Google measure incrementality with $5,000, Fred laid out the measurement revolution. He has argued that Bayesian models ask fundamentally different, and more decision-useful questions compared to traditional frequentist approaches. Rather than asking whether a result is statistically significant, Bayesian methods ask a more practical question: given what we already know, how likely is this to be true?
The practical implication is staggering. Traditional frequentist testing required massive budgets and extended timeframes to prove lift. A $5,000 test split 50/50 with $2 average CPC gives you 1,250 clicks per variant, nowhere near enough for statistical significance under old models. But Bayesian methods, leveraging Google’s informative priors from historical campaign data, can deliver actionable insights from those same numbers.
Fred has explained that Google’s approach doesn’t evaluate tests entirely in isolation. Instead, the system draws on informative priors from large volumes of historical campaign data, hierarchical modeling that groups tests with similar campaigns, and probabilistic outputs that replace p-values with likelihoods.
What this means for practitioners: incrementality testing via Bayesian models sometimes now requires only $5K and seven days to run. That democratizes true lift measurement for mid-market advertisers who could never justify the investment before. The implications cascade through every budget allocation conversation.
Fred has been a prominent voice at industry events exploring these exact questions, consistently emphasizing that measurement isn’t just a nice-to-have anymore. It’s the only way to separate signal from platform-reported noise.
His framework for automation layering, adding intelligent tools and rules on top of Google’s automation rather than fighting it, has become the dominant approach among sophisticated practitioners. You can’t out-algo Google. But you can inform it better than your competitors.
On value-based bidding, Fred has been equally direct in his analysis: Smart Bidding is not set-and-forget. It requires clean conversion data, meaningful value signals, and continuous refinement. The practitioners who treat automation as a starting point rather than an endpoint are the ones seeing results.
Brad Geddes: Keywordless Search and Match Type Reckoning
Brad Geddes, co-founder of Adalysis and one of the longest-tenured voices in paid search, spent last year documenting the quiet revolution happening inside Google’s matching systems.
In his analysis of keywordless Search evolution, Brad revealed uncomfortable truths about how the system operates. Following his investigation, it’s clear that broad match combined with Smart Bidding now treats keyword intent as a suggestion rather than a command. At the same time, reporting changes have made it harder to see what’s actually working.
The problem goes deeper than most realize. As Brad has documented, Google’s AI-powered matching can trump existing exact and phrase match keywords and then claim credit for the conversions and revenue. What appears to be incremental performance from keywordless matching is often cannibalized performance from keywords that were already working.
The implications are profound. Google’s evolution toward AI-powered search creates fundamental attribution problems that prevent advertisers from accurately measuring campaign performance. The system claims credit for conversions that would have occurred through existing exact and phrase match keywords, essentially inflating its own numbers by cannibalizing traffic from more restrictive match types.
Google Ads Product Liaison Ginny Marvin has responded to these concerns by explaining the underlying mechanism. The matching often occurs because of autocomplete suggestions and contextual signals. This represents a shift from standard matching, as Google increasingly determines relevance by inferred intent, similar to how Lens or AI Overviews work, versus just the raw text query. Keywords as we’ve known them are now truly becoming suggestions to an AI system that decides what you really meant.
Brad has noted that search term reports now include a Category column indicating whether terms came from keywordless matching or search themes. This is useful, but it doesn’t solve the core problem: when AI-powered matching decides it knows better than a carefully structured keyword strategy.
His workflow recommendation: add all exact and phrase match keywords as broad match variants to separate performance by match type. Without broad match versions, Google assigns keywordless data to exact or phrase keywords, making accurate performance evaluation impossible. We see greater complexity, not less, but it’s the only way to grasp what’s actually happening.
Throughout 2025, Brad observed accounts in which exact match stopped serving almost entirely, broad match became the primary driver, and phrase match underdelivered relative to historical patterns. The matching changes last year were among the most dramatic in years.
Mike Ryan: PMax Whisperer’s Warning
Mike Ryan, Head of Ecommerce Insights at Smarter Ecommerce (smec) and host of the Growing Ecommerce podcast, has spent three years documenting what actually happens inside Google’s most automated campaign type. AdExchanger has referred to him as The PMax Whisperer for good reason.
In his State of Performance Max analysis of over 4,000 campaigns, Mike identified an uncomfortable paradox. His concern with Performance Max campaigns is that they work too well, a be-careful-what-you-wish-for scenario. When you max performance, the campaign necessarily scales into warm traffic: bestselling products, remarketing, and brand search. That’s because these types of traffic all typically convert really well.
The implication, as Mike has argued, is that the actual job of any given marketing channel is to work hard for your business. The hard work of incremental marketing isn’t always pretty and doesn’t always scale well.
Mike has coined a phrase that captures the strategic moment perfectly: PMax is a new advertising platform incubating within the old one. Other campaign types, he has predicted, will shrink into something vestigial, like a tailbone or an appendix.
On the hybrid approach that many advertisers pursue, Mike has cautioned that running PMax and Shopping in parallel needs to be done carefully. This is especially true since Google announced that PMax no longer takes priority over Shopping. While many advertisers welcomed this news, it comes with the trade-off that these campaign types can cause bids to escalate.
In his late 2025 analysis of what he calls the Clicked vs. Bought Dilemma, Mike exposed a massive technical flaw that most advertisers overlook. In a keywordless world, product targeting is the main lever. But Google places its bids based on the product a user clicks, while being completely blind to which product they actually bought. This discrepancy between click-level targeting and purchase-level reality represents a fundamental misalignment in how PMax optimizes.
Mike’s ongoing assessment of PMax updates reflects a practitioner trying to keep up with a moving target: more control, more transparency, more pro features, kind of. Many of last year’s updates are genuine game changers, long overdue and genuinely useful. Others have left practitioners with more questions than answers.
On ROAS targets and conversion volume, Mike has emphasized that Performance Max campaigns need at least 30 monthly conversions, and ideally 60 or more, for optimal performance. As campaigns advance in number of monthly conversions, target ROAS accuracy improves dramatically while the share of underperforming campaigns dwindles.
His perspective on AI-powered Search raises important questions about whether Google’s newest matching capabilities represent the future or a hidden trap. Analysis of hundreds of e-commerce campaigns has revealed concerning patterns that practitioners need to monitor carefully.
Dennis Goedegebuure: User Experience – The New Land Grab
Dennis Goedegebuure is a strategist focused on where paid, organic, and platform power actually collide — not where Google says they do. Dennis brings a strategic lens shaped by years of working inside large, complex organizations where second-order effects matter. Through senior strategy roles and hands-on involvement in M&A initiatives for companies, he has developed a habit of looking three to four moves ahead—anticipating how platforms, incentives, and market structure evolve over time. That experience informs his perspective today: less focused on short-term optimization, more on how decisions compound once scale, power, and distribution collide.
In a recent essay on Google’s evolving ecommerce layer, he argued that what looks like “improved user experience” is more accurately a structural land grab. The introduction of deeper checkout, product, and merchant layers isn’t neutral plumbing. It’s Google inserting itself further downstream, where intent hardens into transaction, while quietly reframing brands and retailers as interchangeable inventory providers. The promise is convenience. The cost is leverage.
The uncomfortable implication is that brands are being trained to optimize for visibility inside Google’s systems while steadily losing ownership of the relationship that visibility creates. Paid and organic no longer operate as parallel tracks; they’re being merged into a single ecosystem where Google sets the rules, captures the data, and arbitrates the value. Goedegebuure’s take is blunt: if your growth strategy depends on a platform whose incentives diverge from yours, you’re not building an advantage — you’re renting one. You can read the full argument here: https://thenextcorner.net/the-
Kirk Williams: Earning the Strategic Seat
Kirk Williams, Owner of ZATO Marketing, brings a perspective honed in Ads management. Kirk’s take on Performance Max evolution reflects hard-won pragmatism. In industry research on the State of PMax, he offered a balanced assessment: Google deserves credit for addressing many of the major issues that plagued PMax at launch.
While not a blanket endorsement, it’s recognition that fighting the platform’s direction is futile. The question is how to use it strategically. Kirk’s emphasis on small-budget PPC management and strategic testing reflects a core truth: the principles of sound advertising don’t change even when the tools do.
Kirk’s focus on avoiding what he calls PPC frenzy, the temptation to chase every new feature, resonates as platforms pile on capabilities. The practitioners who outperform aren’t the ones using every tool. They’re the ones using the right tools for their specific situation.
On the Google Shopping ecosystem specifically, Kirk has been direct: shopping feeds are now the foundation of e-commerce advertising. Whether running Standard Shopping, PMax, or both, feed quality determines the performance ceiling.
Thomas Eccel: Demand Gen Whisperer
Thomas Eccel, Founder/Managing Director of AdSea Innovations and one of the most influential voices in the PPC community, has carved out expertise in what many consider the next major battleground: Google Demand Gen campaigns.
With over 20 million euros managed in Google Ads spend (with Demand Gen accounting for a significant portion), Thomas has earned the nickname Demand Gen Whisperer for his early mastery of the format. His insights are particularly relevant as Google sunsets Video Action Campaigns and migrates everything to Demand Gen.
On lookalike audiences, exclusive to Demand Gen campaigns, Thomas has identified them as one of the biggest key performance boosters available. The quality of first-party data becomes critical: the more comprehensive and relevant the data, the better Google’s algorithm can identify similar users.
His Full Funnel Domination strategy provides a framework for Demand Gen that most practitioners are still figuring out. Rather than relying on Google’s default automated channel allocation, this approach segments Ad Groups by intent level, optimizing for awareness, consideration, and conversion.
The key insight from Thomas’s work: with Google’s channel controls in Demand Gen, advertisers can now target specific placements at the ad group level, meaning Shorts-only, YouTube in-feed, or Discover-only campaigns are possible. This is a game changer for advertisers looking to refine their strategy and avoid wasted spend.
The billing model change from pure CPM to mixed CPM/CPC represents a fundamental shift in how advertisers should think about Demand Gen costs. Different platforms within the ecosystem have unique engagement metrics: Gmail charges for teaser clicks, YouTube charges per view, Discover video uses an engaged view model. Understanding these nuances separates sophisticated practitioners from those just running campaigns.
On the Video Action Campaign sunset, Thomas has noted the timeline: new Video Action campaigns are no longer available, with existing campaigns migrated into Demand Gen. Practitioners who hadn’t tested Demand Gen by the end of last year are now playing catch-up.
Thomas’s approach to audience seeding shows the cross-platform sophistication required: using Meta’s lookalike data to inform Google Demand Gen targeting, recognizing that first-party data assets don’t belong to any single platform.
Ed Leake: Systems Thinker
Ed Leake, creator of the God Tier Ads Framework, represents a different approach to the automation challenge – one built on systematization rather than constant reinvention. His philosophy cuts against the industry’s obsession with novelty. As Ed has consistently argued, Google hasn’t changed everything. The tools evolve, but the winning strategies don’t. Know your audience. Build the right structure. Optimize what matters.
Ed has captured the automation paradox in his ongoing commentary: practitioners need to shed the delusion that they can out-algo Google. He’s not a proponent of simply handing control of an account over, but the era of trying to beat the algorithm with manual tactics is over.
The framework approach, 400+ step checklists, SOPs, and templates, reflects a maturing industry. When the platforms automate the tactical work, competitive advantage shifts to the systems that wrap around that automation. Not fighting Google, but not surrendering to it either. Creating the infrastructure that makes automation work for specific business goals.
Ed’s insight on differentiation in an automated world is crucial: when everyone’s using the same tech, no one has an advantage. The key to winning has to come from outside the Google Ads account, the offer, the landing page, the data infrastructure.
In practical reality, as Ed has emphasized, a systematized and repeatable process drives consistent results no matter how many features Google changes or hides. The platform will keep changing. The fundamentals of good advertising won’t.
His take on Smart Bidding challenges industry myths: practitioners don’t have to constantly do new things to deliver consistent results. The churn of features and updates can be paralyzing. The practitioners who focus on what actually moves the needle, rather than chasing every beta feature, are the ones building sustainable practices.
Jyll Saskin Gales: Guiding Philosophy
Jyll Saskin Gales, Google Ads Coach, has articulated the philosophical shift that automation demands.
Her framing captures the new reality: guiding, not controlling, is the new marketer’s mantra. The platforms will target. The platforms will bid. The practitioner’s job is to guide those systems toward business goals, not to control every input.
Jyll’s insight about bidding vs. targeting represents a fundamental shift: bidding may now matter more than targeting in automated systems. When targeting happens automatically, the lever advertisers control is what they’re willing to pay for different outcomes. Value-based bidding becomes the primary expression of strategy.
Her approach to AI tools shows how marketers can augment rather than resist and uses Gemini and NotebookLM to scale expertise, generating content and frameworks faster than manual work allows. The AI multiplies its expertise rather than replacing humans.
In her October 2025 Search Engine Land analysis on how audience signals drive PMax, Jyll cut through common confusion about what actually controls Performance Max targeting. The assumptions about how PMax works often don’t match reality. Audience signals don’t function the way many practitioners assume.
On the Power Pack, Google’s trio of AI-powered Search, Performance Max, and Demand Gen, Jyll has noted that this represents where Google is heading. The naming convention confusion creates unnecessary complexity, but the direction is clear. Practitioners who understand this trio and how they interact will have an advantage over those still treating each campaign type in isolation.
Julie Bacchini: Algorithm Drift Problem
Julie Bacchini is President of Neptune Moon and Managing Director of PPC Chat. In PPC Chat discussions throughout 2025, Julie captured the practitioner sentiment well. She has never felt more get-off-my-lawn about PPC than she did last year.
Her warning about algorithm drift has resonated across the community: campaigns that slowly expand into irrelevant search terms because the algorithm thinks it’s being helpful. AI appearing in all the places practitioners don’t really want AI. The platforms are eager to automate, but practitioners are dealing with the unintended consequences.
On trust, Julie has been striking in her assessment: from a practitioner perspective, trust in the platforms is at an all-time low. That’s a remarkable statement from someone who’s spent decades in the industry, and it reflects a broader sentiment that the relationship between advertisers and platforms has become strained.
Julie’s core insight, captured in various industry compilations of hard PPC truths, is that AI can execute, but it cannot strategize. Automation is necessary for modern account management, but the practitioner’s job isn’t moving bids anymore. It’s designing the right layers of automation and guardrails so the algorithms work, not the other way around.
Her take on the death of SKAGs (Single Keyword Ad Groups) reflects the structural shift: SKAGs have run their course. The granular control practitioners used to prize is now a liability. It fragments data and makes it harder for the AI to learn.
The battleground has shifted. As Julie has argued, differentiation is the new battleground. If the offer is weak or the landing page is generic, no amount of bid tweaking will save the campaign. The value add for PPC professionals is shifting from technical setup to business consultancy: fixing the offer, the positioning, and the user experience.
Her observation on match types throughout 2025 is that exact match barely served in some accounts, broad match was surprisingly profitable, phrase match underperformed expectations. The matching landscape last year was among the most volatile in years.
AJ Wilcox: LinkedIn Equation
AJ Wilcox is founder of B2Linked and host of the LinkedIn Ads Show podcast. AJ has consistently argued that LinkedIn has priced itself out of the market for certain types of businesses. But for B2B lead gen with customer lifetime value or deal size greater than about $15,000, LinkedIn Advertising is an absolute no-brainer.
That clarity about when LinkedIn makes sense, and when it doesn’t, reflects the maturity the industry needs. Not every platform is right for every advertiser. The economics has to work.
On LinkedIn’s default settings, AJ has warned that they can quietly drain budgets if left unchecked. The common pitfalls: incorrect geography targeting, audience expansion, low-quality Audience Network traffic, high-cost default bidding. These require active management to avoid.
AJ’s ABM insight highlights a less-discussed problem: ABM campaigns can be inefficient if not optimized, as a few accounts may consume most of the budget. The fix requires breaking up campaigns and managing impressions to ensure fair distribution at the account level.
On LinkedIn’s native industry filters, AJ has noted they are unreliable, often misclassifying companies into incorrect categories. The recommendation: upload custom company lists or use LinkedIn Company Page URLs for more precise targeting.
The emerging features AJ has highlighted show where B2B is heading: predictive audiences that identify entire companies showing interest, Live Event Ads that adapt through the event lifecycle, and Microsoft Designer integration for AI-generated creative directly in Campaign Manager.
His warning on document ads is that they’re great for engagement, but advertisers should be careful about lead quality. A download isn’t the same as intent to purchase.
Andrew Lolk: Consolidation Mandate
Andrew Lolk, founder of SavvyRevenue, has been one of the most vocal advocates for the account structure revolution that automation demands. His warning about over-segmentation is urgent: campaigns that were well-structured under manual bidding are now actively hurting performance under Smart Bidding. The data fragmentation that comes from too many campaigns and ad groups destroys the shared learnings the algorithms need.
The winning structure for 2026 is radically simple, as Andrew has argued. Consolidate until the data proves separation is needed. Split campaigns only when there are business reasons, different bid targets, different promotions, products with different margins that need different ROAS targets.
Andrew’s insight about seasonal inventory illustrates a nuance when, for example, you isolate ski jackets from swimwear so each can optimize on its own performance curve. That’s a strategic reason to separate. High impression volume on a keyword is not.
On the minimum viable structure, Andrew challenges marketers about how many campaigns are needed, alongside the minimum structure required to let the business achieve its objectives. Every additional campaign or ad group is a tax on data density.
His approach to testing has evolved with the platforms. A/B testing, in the traditional sense, is becoming harder because the algorithms are constantly adapting. Practitioners need to think about learning periods, signal quality, and statistical validity in new ways.
Andrew’s client communication philosophy addresses one of the hardest conversations, which is telling a client that their elaborate account structure (the one they paid to build) needs to be torn down. But it’s also the most important conversation.
Shawn Walker & Christine Zirnheld: Platform Reality Check
Shawn Walker of Symphonic Digital and Christine Zirnheld, Director of Lead Gen for Cypress North and co-host of Marketing O’Clock, have been among the most candid voices about the tension between platform incentives and advertiser interests.
Shawn’s warns that Smart Bidding, without strict conversion quality thresholds, chases cheap junk leads because they are the easiest conversions to get. The algorithm optimizes for what it’s told to optimize for, and if conversion tracking doesn’t distinguish between good leads and bad leads, the machine can’t either.
Christine’s observation about AI’s actual utility cuts through the hype. She has found AI to be a go-to tool for smoothing the communication gap between complex PPC work and client understanding. Using it to draft clearer explanations, refine messaging, and spark creative concepts. Not replacing strategic thinking. Augmenting communication.
As Greg Finn has observed in their Marketing O’Clock discussions, the Recommendations tab always suggests raising budgets but never lowering them. That’s the platform reality. Google is a for-profit company. Its objectives don’t always perfectly align with what’s best for advertisers.
The practitioners who understand this, and build systems that account for it, will outperform those who follow recommendations uncritically.
On the weekly cadence of changes, as the Marketing O’Clock team has noted, keeping up with Google’s updates has become a part-time job. Practitioners need a filter for what actually matters versus what’s just noise.
Aaron Levy: Training Gap
Aaron Levy, Evangelist at Optmyzr has identified what may be the industry’s most critical blind spot, which is all about how we train the next generation of practitioners.
In recent PPC Town Hall discussions, Aaron diagnosed a key problem in the fact that training needs to focus on developing people who want to “take the TV apart to see how it works.” Curiosity and first-principles thinking matter more than feature knowledge.
The training gap is more about mindset versus tools and tactics. Too much training focuses on what to do, which buttons to click, which features to use. Not enough focuses on why, the underlying principles that make those actions appropriate in specific contexts.
His leadership philosophy reflects a sense of autonomy within clear frameworks, as opposed to strict rules. The practitioners who understand principles can adapt when the tactics change. Those who only know tactics are constantly playing catch-up.
This has immediate implications for hiring and team development. With this line of thinking, it can be argued that it’s important to know the current Google Ads interface, but it’s even more imperative to be able to analyze problems from first principles and adapt their approach as platforms evolve.
On the agency future, Aaron has been clear that the agencies that survive the next wave of automation will be the ones that invest in developing strategic thinkers, not button-pushers. The button-pushing is likely already automated. The thinking isn’t.
Navah Hopkins: Microsoft Difference
Navah Hopkins, Microsoft Advertising liaison, has brought transparency to cross-platform differences that most practitioners overlook. In her analysis of Microsoft Advertising auction mechanics, Navah disclosed a fundamental difference from Google’s approach – exact match gets priority if it’s present, otherwise Ad Rank determines everything.
Microsoft gives exact match priority over Performance Max. Understanding these platform-specific behaviors is essential for cross-platform advertisers.
On last year’s biggest changes, Navah observed that brand media buys are finally becoming mainstream for performance marketers. That integration of brand and performance, long treated as separate disciplines, is accelerating.
Her concept of PPC insurance, tools and processes that serve as safety nets for automation, frames the practitioner’s role precisely. Focus on creating the backup systems for when automation does something unexpected. Stop fighting automation where it simply makes sense.
In PPC Chat discussions throughout 2025, Navah contended that it’s really hard for match types to continue to be useful when placements evolve, including SERP queries. The matching changes were massive, exact match barely served in some accounts, broad match was surprisingly profitable, phrase match underperformed expectations. The landscape shifted beneath practitioners’ feet.
Her observation from Google liaison conversations reinforces a core truth that even with all the tech changes, core marketing principles still matter. The fundamentals don’t change even when the platforms do.
Greg Finn: Keywordless Reality Check
Greg Finn, Digital Marketer/Partner of Cypress North and co-host of Marketing O’Clock, has been one of the most direct voices about where paid search is heading. His take on AI-powered Search and the keywordless future is pragmatic and clear – Broad Match and PMax work. AI-powered matching will likely work too.
But he’s clear-eyed that keywords aren’t dead just yet, but they will be. Keywordless matching for Search is absolutely a leap in that direction.
Greg’s frustration with naming conventions resonates with practitioners: the confusion between different AI-powered campaign types creates unnecessary complexity. When the products blur together, strategic thinking becomes harder.
His observation about Google’s incentive alignment captures the tension between platform guidance and advertiser interests. He notes that the Recommendations tab always suggests raising budgets but never lowering them. Following recommendations uncritically is dangerous. But ignoring the platform’s direction entirely is also dangerous. The path is somewhere between.
On the pace of change, Greg has noted that 2025 felt like three years compressed into one. The number of significant updates to Google Ads was overwhelming even for people who do this full-time.
Brooke Osmundson: CFO Conversation
Brooke Osmundson, Director of Growth Marketing at Smith Micro Software, has reframed what metrics matter from the brand side. In her recent analysis of PPC KPIs, Brooke delivered a clear message that executives are not impressed by screenshots of green arrows. They want to know whether paid media is adding profit, building pipeline value, and supporting long-term growth.
That’s the conversation shift practitioners need to make. The metrics that matter to CFOs are contribution margin, LTV, pipeline impact, not CTR and ROAS in isolation. Speaking the language of finance opens budget conversations that platform metrics never will.
Her analysis of Google’s Year in Review captured the industry’s current rapid evolution, most notably that automation is no longer something advertisers fear or resist. It’s something they’re learning to wield. The question now is who wields automation most effectively.
On the reporting gap, Brooke has observed that most PPC reports are built for practitioners, but leave out the substance of most interest to executives. The translation work, connecting platform metrics to business outcomes, is where the real value gets communicated.
Sam Tomlinson: Economics Reality
Sam Tomlinson, EVP at Warschawski, has brought an economics-driven perspective to paid media that the industry desperately needed during last year’s volatility. In his April 2025 piece Marketing Through The Tariff Storm, Sam provided the framework that now dictates 2026 margin strategy. When COGS rise, basic math dictates that maintaining current margins requires CAC to come down, prices to rise, or AOV to increase.
When COGS rise, the entire marketing equation changes. CAC targets that worked before become unworkable. Either you find efficiency, raise prices, or accept lower margins.
The April 2025 tariff crisis, which coincided with massive selloffs in U.S. equity markets, forced advertisers to confront questions they’d been avoiding. Sam’s framework became essential reading for practitioners navigating margin-first PPC.
His analysis of what he calls “the next era of marketing” critiques the industry’s obsession with more impressions, more clicks, more spend. Everything is about MORE. The sustainable path is efficiency, though, not scale for its own sake.
The implication for practitioners is that marketing-finance alignment is no longer optional. Understanding how paid media fits into the broader business equation, COGS, contribution margin, LTV, is essential. The CFO is your audience now.
On scenario planning, Sam has emphasized that advertisers who have contingency plans, who know their break-even CAC at different margin levels, navigate economic uncertainty better than those caught flat-footed. The lessons from last year’s tariff storm remain the playbook for 2026.
Duane Brown & Casey Gill: Ground Truth on Spend
Duane Brown, CEO & Head of Strategy of Take Some Risk, and Casey Gill, Strategic AI Product Advisor for Iterate.ai bring ground-level reality to industry discussions about spend stability during last year’s economic uncertainty.
Duane’s observation on spend patterns is telling: almost no one’s spending more money. Some clients might spend more, but on average, precious few are spending more, but virtually no one is dropping spend entirely. When spend is down, it’s more because of profitability numbers, needing to spend less to get profit numbers back where they were.
That’s the reality for many advertisers. Budgets aren’t exploding or collapsing, but margin pressure means finding efficiency becomes existential. The conversation has shifted from how to spend more to how to make every dollar work harder.
Casey highlighted the mobile experience gap as one of the biggest killers of profitability. Not focusing on mobile experiences and conversion rates going into peak season is akin to asking to lose money. The consent modals that work fine on desktop become conversion killers on mobile devices. The landing pages optimized for laptops fail on phones. The best practitioners obsess over these details because they know automation can’t fix fundamentally broken user experiences.
Nils Rooijmans & Wijnand Meijer: Technical Layer
Nils Rooijmans, Google Ads scripts developer and automation specialist, is into vibe coding. In recent PPC Town Hall discussions he chatted up the practice becoming more common. Marketers are inherently using AI tools to build workflows and internal tools, even in the absence of traditional programming skills. Tools like Google AI Studio, Lovable, and V0 allow marketers to move faster with rapid prototyping, generating campaign assets from landing pages, and building custom solutions that would have required developers before.
But Nils has been careful about scope. Vibe coding works for internal tools and experimentation. It’s not a replacement for proper development when things need to be robust and scalable. The key is knowing when AI-assisted coding makes sense and when to hand things off to professionals.
Wijnand Meijer, co-founder and CEO of TrueClicks, represents a different approach to the technical layer: building systematic auditing and monitoring infrastructure that catches what automation misses. TrueClicks now automates checks across accounts.
His philosophy contends that a need exists for an independent third party that evaluates how well accounts are managed. When platforms automate execution, independent validation becomes more important, not less.
The technical layer, scripts, tools, auditing systems, represents where sophisticated practitioners differentiate. Not by doing what the platforms do better, but by building the infrastructure that makes platform automation work for specific business goals.
AIMCLEAR’s Perspective
We’ve spent two decades running campaigns across many platforms and watching the terrain transform in real time. What follows represents our beliefs, informed by this research, validated against our own experience, and aligned with what nearly every expert we’ve studied independently concluded.
Automation Layering Is the Only Path Forward
The platforms will automate, and that automation is accelerating. Fighting this direction is futile and counterproductive. Fred Vallaeys has been preaching his concept for many years. Google, Microsoft, Meta, and LinkedIn are investing billions in AI-driven campaign management. They have more data, more compute horsepower, and more engineers than any advertiser or agency. You will not out-algo them.
However, blind acceptance is equally dangerous. The algorithms optimize for what they’re told to optimize for. They don’t understand your business context, your margin structure, your competitive dynamics, or your strategic priorities. They chase easy conversions because that’s what they’re designed to do.
The winning approach is automation layering, building human intelligence on top of platform automation. This means:
- Layering in your own business rules the algorithms can’t see. Margin thresholds, inventory constraints, brand safety requirements, competitive exclusions. The machines don’t know these things unless you tell them.
- Monitoring systems that catch problems before they become expensive. Algorithm drift happens slowly. Campaigns expand into irrelevant queries, budgets shift to underperforming channels, ROAS calculations diverge from actual profitability. By the time you notice in the platform UI, you’ve wasted money.
- Overriding the built-in capabilities for situations the machines weren’t trained on. New product launches, PR crises, competitive responses, seasonal shifts, economic disruptions. The algorithms learn from history. Novel situations require human judgment.
Automation layering puts you in a position to be smarter than advertisers who let the machines run unsupervised.
Incrementality Measurement Has Been Democratized
Proving true marketing lift is no longer an enterprise-only capability. Bayesian testing methods, combined with new measurement frameworks, make it possible to run meaningful incrementality tests with budgets that mid-market advertisers can actually afford.
The old excuse, we can’t afford proper testing, no longer holds. What’s changed:
- Lower budget thresholds. Tests that required $100K+ now produce actionable insights at $5K.
- Shorter timeframes. Campaigns that needed six weeks can now show lift in seven days.
- Probabilistic outputs instead of binary pass/fail. Instead of asking if something is statistically significant, we can ask if something is actually true, given everything we know.
Every serious advertiser should be running incrementality tests, and those who don’t are operating on faith rather than evidence. The conversation has to shift from “trust us, it’s working” to “here’s the probability this will drive real lift.”
The Keywordless Future Is Arriving
Keywords are becoming suggestions rather than commands. AI-powered Search, Performance Max, and the broader shift toward intent-based matching mean the platforms increasingly decide what you meant, not what you typed.
The platforms have explicitly stated they’re determining relevance by inferred intent versus just the raw text query. The matching changes last year were dramatic: exact match barely served in some accounts, broad match outperformed expectations, phrase match underdelivered.
Practitioners who cling to exact match as a security blanket will underperform. The winning approach is:
- Feed the intent signals properly. First-party data, customer lists, purchase history, CRM integration. The more context you provide, the better the algorithms understand your actual customer.
- Accept broader matching and focus on guardrails. Negative keywords, placement exclusions, audience signals. Control what you can control.
- Measure at the business outcome level. Query-level optimization is becoming impossible. Campaign-level and account-level optimization against real business outcomes is the new game.
Keywords will exist for years to come, but their role is diminishing, and practitioners who understand this will adapt their workflows accordingly.
First-Party Data Is Now Existential
The decisive competitive advantage in paid media is no longer tactical excellence inside the platforms. It’s the data you bring to them.
When everyone has access to the same automated tools, the same AI-powered bidding, the same broad match algorithms, the differentiator is what you feed those systems. Brands with strong first-party data infrastructure will consistently outperform those relying on platform defaults.
What this means practically:
- Customer match lists matter more than keyword lists. Your CRM is now a paid media asset.
- Conversion infrastructure marks a competitive advantage. If your tracking captures purchase value, customer LTV indicators, and lead quality signals, your algorithms learn faster and optimize better than competitors with basic tracking.
- Clean data pipelines beat dirty data volume. Garbage in, garbage out. The brands that invest in data hygiene will see compounding returns as the algorithms learn from better signals.
- Lookalike audiences require quality seeds. The better your first-party data, the better the platforms can find similar users. This is especially critical for Demand Gen campaigns where lookalikes are a primary lever.
The gap between data-rich and data-poor advertisers will widen dramatically as automation matures.
Marketing-Finance Alignment Is Mandatory
PPC can no longer operate in a silo. Economic volatility, margin compression, COGS fluctuation, macro uncertainty, means marketing metrics must connect to business metrics.
The math shows that when COGS rise, CAC targets that worked before become unworkable. When margins compress, efficiency becomes existential. When macro uncertainty persists, flexibility in budget allocation matters more than annual plans.
Practitioners who can speak the language of the CFO will have more credibility and more budget than those who only speak platform metrics:
- CAC in context of COGS and contribution margin. Not CAC in isolation.
- ROAS as input, not outcome. What matters is profitable revenue, not attributed revenue.
- LTV-informed bidding and budgeting. Not all customers are equal. Your paid media should know this.
The practitioners who make this translation, who connect marketing activity to business outcomes in language finance execs understand, will be the ones who survive budget cuts and earn expansion.
Marginal Thinking Beats Average Thinking
The industry’s obsession with average ROAS is fundamentally misleading. What matters is marginal ROAS, the return on the last dollar spent, not the blended average.
High-performing campaigns often have excellent average ROAS because they’re spending too little. The incremental opportunity is being left on the table. Underperforming campaigns often look fine on average but are bleeding money at the margin.
Practitioners need to think like economists:
- What happens if we spend the next $1,000? Focus less on what a campaign returned on average.
- Where is the diminishing return curve? Every campaign has one. Finding it is the job.
- Incremental contribution is more important than attributed contribution. The platform will take credit for everything it can. Your job is to find the true lift.
This is uncomfortable thinking for an industry trained on ROAS dashboards. But it’s the only thinking that actually optimizes for business outcomes.
Account Structure Must Simplify
Over-segmentation, the legacy of manual bidding eras, is now actively harmful. Campaigns and ad groups that made sense when humans set bids are fragmenting the data that algorithms need to learn.
The winning structure for 2026:
- Consolidate until the data proves you need to separate.
- Split campaigns for business reasons only. Different margin targets, different seasonality, different competitive dynamics.
- Let the algorithms have signal density. They need data to learn. Starving them with fragmented structures hurts performance.
The practitioners who spent years building elaborate account structures now need to tear them down. The machines need what they need. Give it to them.
Demand Gen Is the Next Battleground
The migration from Video Action Campaigns to Demand Gen represents the next major shift in how practitioners allocate spend. The full-funnel capabilities, lookalike audiences, and channel controls make this campaign type strategic in ways previous video formats weren’t.
What matters in Demand Gen:
- First-party data quality determines lookalike effectiveness. The seed list matters more than the algorithm.
- Channel controls require deliberate strategy. Shorts-only, YouTube in-feed, Discover-only: these are different contexts requiring different approaches, not interchangeable placements.
- The mixed billing model demands new measurement frameworks. CPM plus CPC means different ROI math than pure video campaigns.
Practitioners who master Demand Gen now, before competitors catch up, will have a structural advantage.
Platform Trust Must Be Verified
Trust in platform recommendations should be low and declining. The platforms are optimizing for their interests, which don’t perfectly align with advertiser interests.
- Recommendations consistently suggest spending more, never less.
- Attribution benefits the platform, not the advertiser.
- Automated features expand reach in ways that help platform revenue.
- Default settings often favor platform economics over advertiser economics.
Independent verification is essential. Third-party tools, independent audits, incrementality testing that doesn’t rely on platform data alone. Trust, but verify. The practitioners who accept platform guidance uncritically will underperform those who maintain healthy skepticism.
Mobile Excellence Is Non-Negotiable
Fundamentally broken user experiences cannot be fixed by better advertising. And mobile experiences remain broken at a staggering rate.
Consent models that work on desktop become conversion killers on phones. Landing pages optimized for laptops fail on mobile. Checkout flows designed for the use of a mouse fail for thumbs. Page speed that’s acceptable on fiber becomes unacceptable on cellular.
Obsessing over these details is mandatory because automation can’t fix them. The algorithms will find users and serve ads. What happens after the click is your job. And for most advertisers, mobile is where that experience falls apart.
The Clicked-vs-Bought Problem Is Real
The discrepancy between the product a user clicks and the product they actually buy represents a fundamental challenge in e-commerce advertising that most practitioners ignore.
When Performance Max bids based on clicked products but revenue comes from different products, the optimization signal is misaligned. This creates systematic inefficiency that compounds over time.
Practitioners need to:
- Understand the gap between click attribution and purchase attribution in their accounts.
- Build measurement frameworks that capture cross-product purchasing behavior.
- Advocate for platform improvements that address this misalignment.
The platforms are optimizing for clicks. Your business depends on purchases. That gap matters.
The Practitioner’s Role Is Evolving, Not Disappearing
Every wave of automation triggers predictions that humans will be replaced. It hasn’t happened yet. It won’t happen now.
What’s changing is what humans do. Less time on bid adjustments, keyword mining, and match type management. More time on:
- What are we actually trying to accomplish, and how does paid media serve that goal?
- Creative direction. The machines can optimize delivery. They can’t create compelling messages.
- Measurement architecture. What are we tracking, how are we attributing, and does this reflect actual business outcomes?
- Business integration. How does paid media connect to the rest of the marketing mix and the broader business strategy?
- Exception handling. When the algorithms do something unexpected, humans need to understand why and decide whether to override.
The practitioners who adapt will be more valuable than ever. The practitioners who were defined by tactical execution will struggle. The practitioners who can think strategically and manage automated systems will thrive.
Training Must Evolve
The industry’s approach to developing talent is fundamentally misaligned with what the job now requires. Too much training focuses on what to do, which buttons to click, which features to use. Not enough focus is on why, the underlying principles that make those actions appropriate in specific contexts. Our clients who understand principles can adapt when the tactics change. Those who only know tactics are constantly playing catch-up.
This has immediate implications:
- Hire an agency for thinking as a layer on top of its platform knowledge. The interface will change. The ability to reason from first principles won’t.
- Develop strategic capacity, not just tactical skill. The tactical work is being automated.
- Build business acumen alongside platform expertise. The CFO conversation matters more than ever.
The Insanity Is Over
2025 was the year of anxiety. Marketers saw AI-powered Search launching, Performance Max opacity, match type chaos, and economic uncertainty. The industry spent the year worried about what was happening and what it meant. 2026 is the year of getting shit done. The tools exist and visibility has improved. We know what to do.
Clients who move from anxiety to action will capture disproportionate value. The playbooks are being rewritten in real time. The winners will be those who write them.
The platforms are finally returning visibility into their automated systems. Performance Max got more transparent. AI-powered Search launched with reporting that mostly makes sense. Incrementality testing became more accessible and demand gen emerged as a serious campaign type with real controls.
Remember that automation layering beats automation resistance. Incrementality beats attribution. First-party data beats platform defaults. Finance alignment beats marketing silos. Marginal thinking beats average thinking. Simplified structures beat fragmented ones. Demand Gen mastery becomes differentiating.
Clients who internalize these shifts and build the systems to operationalize them will outperform, but those who fight the direction of travel will struggle. Platforms are always moving. The question is whether you move with them, ahead of them, or against them.
In Part 3 of this series, we’ll explore how paid and organic are merging, not just in theory, but in practice. Brand authority now determines both AI citation and paid algorithmic performance. First-party data bridges the divide. The silos were always artificial. The practitioners who understand that will build unified discovery systems that compound across every surface.
