2025 marked a significant first for me: I was invited to join a select group of 15 PPC thought leaders for the Google Marketing Live (GML) presentation in Mountain View. Along with attending the live event, I gained exclusive early access to announcements two weeks in advance and had the invaluable opportunity to directly engage with the product managers behind these new features.
This privileged access provided me with:
- Extended time to analyze the new features thoroughly.
- Deeper insights beyond what was publicly announced.
- Hands-on experience with features during their beta phase.
In this article, I’ll share my expert perspective on the GML announcements. For those who couldn’t attend or watch the entire presentation, I’ll begin with a concise summary. (And for those concerned about my partner, Filip, he was invited to a livestream of the presentation in Dublin, so he wasn’t left out!)
Key Announcements from Google Marketing Live 2025
Based on my analysis, the most impactful features announced are (in prioritized order):
- Incrementality Testing Enhancements
- Smart Bidding Exploration
- Attributed Brand Searches
- Shoppable CTV (Connected TV)
Advancements in Smart Bidding & Search Max
- AI Max for Search Campaigns: Rolling out in open beta post-GML, this one-click solution integrates AI-powered targeting and creative enhancements into existing campaign setups without disruption. Google reports it typically delivers 14% more conversions at a similar CPA/ROAS, with potential increases up to 27% for campaigns not yet fully optimized with broad match.
- Smart Bidding Exploration: This feature leverages Google AI to bid strategically on searches previously considered out of reach. Hailed as the most significant leap in bidding technology in over a decade, it has demonstrated an 18% increase in unique search query categories generating conversions and a 19% uplift in conversion value.
- Performance Max Updates:
- Channel Performance Reporting: Provides enhanced visibility into how campaigns perform across Google’s various channels, complete with visualization tools and diagnostic capabilities.
- Search Term Reports: Offers the same granular search term data and functionalities previously exclusive to search and shopping campaigns, now available for Performance Max.
Optimizing Tracking and the Click-to-Purchase Journey
- Shoppable CTV: Enables viewers to purchase products directly from their TV screens or seamlessly send items to their phones for continued browsing. This feature will become available for Demand Gen and Performance Max campaigns later this year.
- Shoppable Masthead: Transforms YouTube’s most prominent ad placement on mobile into a shoppable experience, providing a direct path from inspiration to discovery.
- Attributed Brand Searches: Reveals how YouTube advertising influences subsequent brand-related searches on Google. It includes configurable attribution windows (defaulting to 7 days, extendable up to 30 days).
Improvements in Measurement and Data Integration
- Expanded Incrementality Testing: The minimum budget requirement for incrementality testing has been drastically reduced from $100K to $5K. A new Bayesian methodology increases the likelihood of achieving conclusive results. This is now available for all campaign types, with results directly accessible in the Google Ads UI.
- Confidential Matching: Google is expanding this secure data processing feature to the Google Tag Gateway, enhanced conversions, and user-provided data within Google Analytics.
- Data Manager: This new tool serves as a centralized hub for connecting online and offline data sources, offering personalized guidance on data setup. It’s slated for release in Search Ads 360 soon and Campaign Manager 360 later this year, featuring consolidated APIs.
My Takeaways: Reading Between the Lines
A notable shift appears to be underway: Google seems to be moving away from its aggressive “PMax crusade.” We’ve all experienced the relentless push towards Performance Max, with representatives often suggesting Standard Shopping would be sunsetted and PMax would become the sole campaign type. Previously, new feature releases were often exclusive to PMax. What we’ve seen in the announcements the last month is a renewed interest in the entire suite of campaigns inside Google Ads — not just PMax. This echoes the sentiment that I hear when I speak to people inside Google.
Furthermore, there was a surprising scarcity of truly “big” or unexpected announcements. Many of the highlighted features have been accessible in beta programs for the past 6-12 months. Perhaps this is a testament to the value of gaining early beta access, but I found most of the GML announcements to be already familiar.
My Actionable Strategies: Putting the New GML Features to Work
While there were a lot of announcements, four features really caught my eye for immediate use, plus a bonus one:
- Incrementality Testing
- Smart Bidding Exploration
- AI Max for Search
- Attributed Brand Searches
- Bonus: Shoppable CTV
(I’ll be candid: features like Confidential Matching, while technically interesting, tend to make my eyes glaze over, and I doubt many of our clients will allocate budgets for YouTube.com mastheads.)
So, here’s how I’m planning to use these key announcements:
1. Incrementality Testing: Unlocking True Campaign Value
This enhanced feature reminds me of Google Ads Experiments, a native A/B testing tool that is, frankly, criminally underutilized. The ability to test the impact of various changes—from targets to bidding strategies—is invaluable. I predict that, like Experiments, built-in incrementality testing will be an incredibly powerful tool that, unfortunately, less than 1% of Google Ads professionals will actively use. This is a missed opportunity.
I see five main ways I’ll use incrementality testing:
- Brand Bidding Efficacy: Finally, a robust way to settle the perennial debate about the true incremental value of brand bidding in many accounts.
- Search vs. Shopping Dynamics: Shopping campaigns are dominant for our e-commerce advertisers. But what is the precise contribution of Search? How much incremental value do we gain by running Search alongside Shopping? Does a user seeing both a Shopping and a Search ad for the same query result in a 1+1=2, 1.8, or even 3 effect?
- Retargeting Effectiveness: I’ve long been skeptical of Google’s Display retargeting, often seeing low volume and underwhelming ROAS, even for high-traffic advertisers. While it’s not a standard part of our frameworks, I’m eager to conduct more rigorous testing.
Aggressive Bidding Impact (e.g., Black Friday): This isn’t just about standard bidding changes. For peak periods like Black Friday, I’m keen to test whether our typically aggressive bidding strategies deliver the expected incremental returns.- Video Advertising’s Halo Effect: What is the quantifiable impact of a scaled YouTube program on Shopping Ad performance? We know brand-building activities can lift performance, but by how much? Is it enough to offset the often lower (reported) ROAS from YouTube itself?
2. Smart Bidding Exploration (SBE): Venturing into New Conversion Territories
Having experimented with SBE during its beta, I’m convinced of its significant potential. For a long time, I’ve observed that simply lowering ROAS targets on Smart Bidding didn’t always achieve the desired outcome of exploration; it often just increased bids for existing placements. Smart Bidding Exploration directly addresses this by actively seeking new traffic sources.
Google’s claim that this represents the “biggest leap in Smart Bidding technology in over 10 years” is a bold statement, especially considering the substantial improvements to Smart Bidding 4-5 years ago. This makes SBE a particularly compelling feature.
Here are three ways I plan to use SBE:
- During “Peacetime” Operations: I categorize account management into “Peacetime” (targets are being met, allowing for investment in scaling tests) and “Wartime” (focus is on combating inefficient spend). I anticipate routinely running SBE during Peacetime, perhaps even continuously, to uncover new growth opportunities.
- Pre-Peak Season Ramp-Up: In the weeks leading up to and the initial stages of a peak season, I will deploy SBE aggressively. The goal is to identify search terms that were previously excluded or haven’t been explored sufficiently in the current year.
- Mid-Peak Season Optimization: If a peak season is performing well (indicated by high ROAS), I will continue to leverage SBE aggressively to maximize conversion volume.
Conversely, I would strongly advise against using SBE if you are struggling to meet ROAS targets in your account, as it could introduce additional volatility.
3. AI Max for Search: Enhancing Scale and Insights
I see AI Max for Search as being like integrating Dynamic Search Ad functionality directly into existing ad groups. Overall, I believe it’s a positive development. The idea of driving more spend through Search campaigns is appealing to me, as aggregating more data can provide valuable insights for crafting more effective ad copy.
I expect to use AI Max mainly in two situations:
- Accounts with Large Inventories: This is a natural fit, similar to our current use of Dynamic Search Ads. I foresee transitioning many of our Search campaigns to AI Max, pending successful trials.
- Expanding Scale When ROAS is Strong: When Search performance has plateaued but ROAS margins are healthy, we often turn to broad match to scale. AI Max could become our preferred method in such situations.
However, there are still unknowns with AI Max. Just as with using Dynamic Search Ads within existing ad groups (which I generally oppose due to the potential for messy, irrelevant traffic), AI Max might inadvertently expand to too many irrelevant search terms if not carefully managed.
4. Attributed Brand Searches (YouTube Only): Quantifying YouTube’s Impact
One of the ongoing headaches with YouTube Ads is getting solid data to demonstrate their true value. While we often observe an increase in brand searches when running aggressive YouTube campaigns, it’s always been difficult to determine how many of those searches would have occurred organically.
Attributed Brand Searches helps fix this by letting us define our brand terms and then clearly see the volume of additional brand searches generated by YouTube advertising.
5. Bonus: Shoppable CTV (YouTube Only): The Evolving Landscape of TV Advertising
I’ve added Shoppable CTV here because Connected TV (as opposed to traditional linear/cable TV) is an interesting new area in advertising. With up to 50% of YouTube content now consumed via TV screens, it’s crucial to consider how users interact with ads beyond mobile and desktop.
Amazon’s Prime Video already features shoppable ads within its content, so YouTube’s adoption of this technology was inevitable.
While I haven’t personally tested this feature yet, it will be fascinating to see how it impacts the ability to track and attribute conversions from YouTube ads displayed on television.
Key Questions Posed to Google Product Managers
During our briefing, we asked some key questions. Here’s what we learned from their answers:
- Q: How reliable is Incrementality Testing, especially with reduced data levels?
- A: Google has implemented a new Bayesian methodology designed to achieve conclusive results even with less data. The significant reduction in the minimum budget requirement (from $100K to $5K) makes this powerful tool more accessible without compromising reliability.
Regarding Smart Bidding Exploration:
- Q: How does lowering ROAS targets differ fundamentally from Smart Bidding Exploration?
- A: Lowering ROAS targets primarily increases bids across existing matched keywords without necessarily expanding traffic diversity. In contrast, Smart Bidding Exploration actively seeks new traffic sources and untapped query categories to capture additional conversion value.
- Q: Will Smart Bidding Exploration be compatible with Portfolio Bid Strategies?
- A: Yes, it is already available for portfolios in the closed beta and is scheduled for inclusion in the open beta following GML.
- Q: Can we effectively track the performance of exploration efforts in reporting?
- A: Yes, the bid strategy report will provide data on unique search categories, clicks, and conversions specifically attributable to exploration efforts.
Regarding Attributed Brand Searches:
- Q: What are the setup requirements for Attributed Brand Searches?
- A: The primary requirement is to set up brand mapping, which involves specifying the brand names you want to track. Once configured, it becomes available as a new metric within Google Ads, utilizing a default 7-day attribution window that can be adjusted as needed.
My Rank of the Announcements
Wrapping Up: So, When Can We Actually Use These Features?
Alright, let’s talk about the elephant in the room: the waiting game. This is often the most frustrating part of Google Marketing Live. We see all these new tools, but getting our hands on them isn’t always immediate.
Many of the announcements are currently in closed betas. That means folks like me get an early look, but they’re likely not taking new advertisers for these right now. For everything else, Google’s rollout timelines can be a bit of a moving target. I did my best to jot down every mention of “access,” but here’s the general gist: expect most of these new features to hit open beta in the weeks following the GML event.
Just to clarify, open beta usually means you can request access, often through your Google Rep, and you’ll likely be granted it.
The catch? Yes, you’ll probably need that Google Rep to open the door for many of these, which I know can be a hurdle for some. My advice? Start having those conversations with your reps now. Get on their radar, express your interest in specific features, and begin planning how you’ll test them once they’re available.
